Global, regional and national burden of glaucoma from 1990 to 2021 and projections to 2050: a retrospective cross-sectional study

Abstract

Objectives

This study assessed the global burden of glaucoma using data from the Global Burden of Disease (GBD) 2021 study. The analysis of epidemiological trends aimed to inform future public health prevention strategies.

Design

Retrospective cross-sectional study.

Participants

None.

Methods

Analysis of 1990–2021 GBD data on glaucoma prevalence, disability-adjusted life years (DALYs), age-standardised prevalence rates (ASPR), and age-standardised DALY rates (ASDR). Estimated annual percentage changes (EAPC) were calculated, Joinpoint regression identified trend changes, and Autoregressive Integrated Moving Average (ARIMA) modelling projected the burden for the year 2050.

Results

Globally, the number of prevalent glaucoma cases increased from 4 072 106.59 (95% uncertainty interval (UI) 3 489 888.7 to 4 752 867.3) in 1990 to 7 587 672.9 (95% UI 6 522 906 to 8 917 725.4) in 2021. Concurrently, DALYs increased from 467 600.4 (95% UI 323 490.5 to 648 641.6) in 1990 to 759 900.2 (95% UI 530 942.9 to 1 049 127.2) in 2021. In contrast, the ASPR and ASDR declined to 90.1 per 100 000 population (95% UI 77.8 to 105.5) and 9.1 per 100 000 population (95% UI 6.3 to 12.5) in 2021, respectively. During the COVID-19 pandemic period (2019–2021), the slowest growth rates in crude case numbers and overall disease burden were observed, accompanied by the most pronounced decline in annual percentage change of ASPR. The highest estimates for both case counts and DALYs were identified in the 70–74 age group, with males demonstrating higher prevalence rates than females. Furthermore, regions with lower Sociodemographic Index (SDI) values bore a disproportionately higher burden of glaucoma.

Conclusion

These findings underscore the need to strengthen early screening and treatment of glaucoma, particularly in ageing populations, male groups and low SDI regions. We urge cautious interpretation of COVID-19 related data and vigilance against potential post-pandemic surges in burden. Critical strategies include enhanced screening and intervention for high-risk groups, targeted prevention measures and integration of ophthalmic care into public health emergency frameworks to alleviate the disease burden.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The application of advanced statistical approaches, including Joinpoint regression, decomposition analysis and Autoregressive Integrated Moving Average modelling, to the Global Burden of Disease 2021 data enhances the detection of trends and the robustness of forecasts, while the inherent reliance on modelled estimates may introduce residual uncertainty.

• Age-standardised metrics and Sociodemographic Index stratification improve cross-regional interpretability, yet variations in local healthcare reporting quality could affect data granularity.

• Explicit evaluation of COVID-19’s potential impact on glaucoma burden strengthens temporal analysis, though indirect effects remain challenging to quantify fully.

• While the focus on the population aged 45 years and older is consistent with the typical epidemiology of glaucoma, it consequently limits insights into early-onset cases and may lead to an underestimation of the condition’s lifetime burden.

Introduction

Glaucoma is a blinding ocular disorder characterised by progressive apoptosis of retinal ganglion cells, optic nerve atrophy and visual field defects, which can occur with or without elevated intraocular pressure (IOP). It ranks among the leading causes of irreversible blindness.¹ A significant subset of patients remains asymptomatic in early stages, often leading to delayed diagnosis until the disease has progressed to an advanced state with irreversible visual impairment,^{2 3} which severely compromises quality of life. Currently, approximately 95 million people worldwide are affected by glaucoma, with projections indicating a rise to 111.8 million cases by 2040.⁴ The Global Burden of Disease (GBD) Study and the Vision Loss Expert Group of the GBD 2019 Blindness and Vision Impairment Collaborators reported that in 2020, glaucoma caused blindness in 3.61 million people worldwide, with an additional 4.14 million individuals experiencing moderate to severe vision impairment attributable to this condition. Glaucoma accounted for 8.39% of all global blindness cases that year.⁵ A cost-effectiveness analysis of glaucoma in the US revealed that the mean healthcare expenditure over a 5-year horizon ranges from US$ 6172 to US$ 10 075 per patient.⁶ In China, the annual economic burden for individuals with unilateral and bilateral blindness due to glaucoma is estimated at between US$ 1080 and 8920.⁷

Glaucoma, a chronic and progressive optic neuropathy with a long disease duration and high risk of blindness, imposes substantial socioeconomic burdens worldwide. Although early intervention can effectively slow vision loss, the global rate of underdiagnosis remains alarmingly high, exceeding 90% in some low- and middle-income countries compared with approximately 50% in high-income nations.⁸ Furthermore, glaucoma prevalence exhibits significant variation across racial and ethnic groups,^{9 10} underscoring the need for standardised global assessment frameworks to enhance detection and treatment. The GBD Study, widely regarded as a gold standard for disease burden evaluation, provides critical evidence for public health policymaking. Recent analyses demonstrate the increasing prevalence of glaucoma and uneven geographic distribution by age and sex.11,13 However, evolving lifestyles, increasing life expectancy and major global health crises are likely reshaping its epidemiological trends. Despite these dynamic influences, contemporary large-scale epidemiological investigations remain limited, highlighting a clear demand for updated population-level research.

Utilising GBD 2021 data, this study advances prior research by applying an integrated analytical framework that combines Joinpoint regression, decomposition analysis and Autoregressive Integrated Moving Average (ARIMA) modelling. Our multi-scale approach enabled a comprehensive assessment of the glaucoma burden across geographic levels, complemented by future projections to provide updated evidence for health policymaking. The COVID-19 pandemic (2019–2021) disrupted healthcare systems globally, with potential implications for exacerbating health disparities.¹⁴ Contextual factors such as overwhelmed medical infrastructure, prioritisation of severe acute conditions and heterogeneous clinical protocols may have differentially impacted glaucoma management,¹⁵ thereby potentially influencing the GBD 2021 estimates. Our analysis therefore incorporated a specific evaluation of these pandemic-related confounders to ensure a robust interpretation of the observed burden trends.

Methods

Overview

The GBD 2021 study, produced by the Institute for Health Metrics and Evaluation, provides the most comprehensive and up-to-date assessment of global health. It estimates key epidemiological metrics, such as incidence, prevalence, mortality and disability-adjusted life years (DALYs), for 371 diseases and injuries across 204 countries and territories, stratified by 23 age groups. This latest iteration comprehensively revises all estimates from 1990 to 2021, thereby establishing new benchmarks for global health comparisons.¹⁶ As a retrospective cross-sectional analysis of fully de-identified data from the GBD study, this research adheres to the principles of the Declaration of Helsinki and is exempt from Institutional Review Board approval and informed consent requirements. The reporting of this study follows the Guidelines for Accurate and Transparent Health Estimates Reporting.¹⁷

Data sources

The GBD 2021 database synthesises a wide array of sources, including scientific literature, surveys, registries, and administrative and clinical data, to generate comprehensive global estimates. Its construction incorporated 100 983 distinct data sources, of which 75 459 were dedicated to estimating non-fatal outcomes such as glaucoma. To ensure robustness, the GBD framework employs cross-validation among source types and leverages collaborative networks to enhance data representation, particularly in high-burden regions such as Sub-Saharan Africa. This systematic approach aims to minimise source-specific bias and support reliable burden estimation across 204 countries and territories.¹⁶

Glaucoma prevalence and DALYs were estimated using the GBD 2021 database, with case definitions corresponding to ICD-10 codes H40-H40.6 and H40.8-H40.9. The estimates were generated through DisMod-MR 2.1, which is a Bayesian meta-regression tool designed to systematically integrate diverse epidemiological sources while adjusting for systematic variations in study methodologies and diagnostic criteria. This modelling approach employs a hierarchical structure that progresses from global to super-region, region and country levels, which ensures internal consistency and yields reliable estimates for populations with limited primary data. The GBD study frequently employs ensemble modelling techniques to determine optimal estimates across multiple model specifications. All quantitative results are reported with 95% uncertainty interval (UI) derived from 500 computational iterations, thereby comprehensively capturing uncertainty from sampling variability, model specification and covariate estimation. Age-standardised prevalence rates (ASPR) and age-standardised DALY rates (ASDR) per 100 000 population were calculated to enable standardised cross-regional comparisons.¹⁶ Sociodemographic development was assessed using the Sociodemographic Index (SDI), combining income, education and fertility rates. Countries and regions were categorised into five SDI tiers: low, low-middle, middle, high-middle and high.

Patient and public involvement

None.

Joinpoint regression analysis

The Joinpoint regression approach assesses temporal trends in glaucoma-attributable disease burden metrics by identifying statistically significant inflection points (joinpoints) through sequential linear models. It offers better data fitting than single-trend models.¹⁸ Specifically, we adopted a logarithmic linear model (ln y = β*x) for segmented regression. To determine the optimal number of join points, we first used the grid search method (GSM) to generate all possible join point combinations, calculated the mean squared error (MSE) for each scenario, and selected the combination with the smallest MSE as the initial candidate. The final number of join points was validated via Monte Carlo permutation tests, with the model restricting the number of potential join points to a minimum of 0 and a maximum of 5. The test iteratively compared models with an increasing number of join points (from k=0 to k=5) until the optimal model was identified. Based on the minimum MSE from the GSM and the validation results of Monte Carlo permutation tests, the optimal model for our analysis included five joinpoints. For trend quantification, we calculated the annual percentage changes (APC) for the period 1990–2021. The APC was derived using the formula: APC=(eˆβ − 1)×100%, where β is the regression coefficient of the logarithmic linear model. An APC greater than 0 indicates an upward trend within a segment, while an APC less than 0 indicates a downward trend. All analyses were performed using Joinpoint Regression Software V.5.1.0 (National Cancer Institute).

ARIMA model for time series forecasting

The ARIMA model represents a classical time series forecasting methodology that integrates autoregressive (AR) and moving average (MA) components with differencing (d) operations to achieve stationarity. In the ARIMA (p,d,q) model, p represents the AR order, d is the differencing degree needed to achieve stationarity, and q denotes the MA order. Through parameter optimisation, the ARIMA model demonstrates enhanced capability in handling complex datasets and improving predictive accuracy. Model specification followed a structured diagnostic procedure. Stationarity was assessed using the Augmented Dickey-Fuller test, while optimal parameter combinations were identified through autocorrelation function and partial autocorrelation function analysis, with final model selection guided by information-theoretic criteria. Model adequacy was verified through residual diagnostics, including the Ljung-Box Q-testing, to confirm independent and normally distributed residuals.¹⁹ This implementation adheres to Box-Jenkins methodology, ensuring model parsimony and maintaining predictive accuracy for epidemiological time series analysis.

Statistical analysis

This study analysed the burden of glaucoma exclusively in populations aged 45 years and older from 1990 to 2021, as data for younger age groups were unavailable. We calculated age-standardised rates (ASR) with 95% UIs for prevalence and DALYs across demographic and geographic strata, assessing temporal trends using estimated annual percentage changes (EAPC). Joinpoint regression identified significant inflection points in these trends, while decomposition analysis quantified the respective contributions of population growth, population ageing and changes in epidemiological rates. Future burden through 2050 was projected using ARIMA modelling. All analyses were performed in R V.4.4.2, employing a statistical significance threshold of p<0.05.

Results

Global burden of glaucoma and time trends: a comprehensive analysis from 1990 to 2021

Table 1 and online supplemental table 1 present the global burden of glaucoma from 1990 to 2021. Globally, the number of prevalent cases increased by 86.3% to 7 587 672.9 in 2021, while the ASPR declined to 90.1 per 100 000. Similarly, DALYs increased by 62.5% to 759 900.2, whereas the ASDR declined to 9.1 per 100 000. Middle SDI regions recorded the highest number of prevalent cases (2 611 142.1 cases in 2021), with Western Sub-Saharan Africa showing the highest ASR (ASPR: 294.4 per 100 000; ASDR: 32.5 per 100 000). Conversely, Central Europe reported the lowest burden (ASPR: 38.6 per 100 000). All 27 GBD regions exhibited declining ASR, with South Asia demonstrating the most significant reductions.

Table 1. Glaucoma prevalence and ASPR in 1990 and 2021 with 1990–2021 temporal trends.

Prevalence location	1990		2021		1990–2021 EAPC in ASPR (95% CI)
	Prevalence counts (95% UI)	ASPR per 100 000 (95% UI)	Prevalence counts (95% UI)	ASPR per 100 000 (95% UI)
Global	4 072 106.5 (3 489 888.7 to 4 752 867.3)	116.3 (100.8 to 136.3)	7 587 672.9 (6 522 906 to 8 917 725.4)	90.1 (77.8 to 105.5)	−0.73 (−0.78 to −0.69)
High SDI	646 582.1 (556 759.6 to 753 570.3)	57.8 (49.9 to 67)	1 191 263.2 (1 031 211.9 to 1 391 418.5)	49.1 (42.6 to 57.1)	−0.48 (−0.56 to −0.41)
High-middle SDI	889 767.3 (762 658.6 to 1 047 424.1)	102.8 (88.6 to 120.5)	1 400 629.5 (1 205 785.4 to 1 634 878.9)	71 (61.2 to 82.8)	−0.92 (−1.01 to −0.84)
Middle SDI	1 213 358.3 (1 024 437.9 to 1 414 660.6)	149.2 (128.4 to 176.3)	2 611 142.1 (2 238 156.3 to 3 074 011.7)	105.5 (90.9 to 123.6)	−0.9 (−0.98 to −0.83)
Low-middle SDI	897 512.5 (755 752.4 to 1 059 114.5)	184 (156.5 to 216.9)	1 617 523.2 (1 366 423.4 to 1 913 844.3)	129.2 (109.8 to 153.2)	−1.29 (−1.35 to −1.23)
Low SDI	421 146.7 (355 337.7 to 495 268.9)	245 (209.6 to 287.8)	760 939.2 (643 085.6 to 898 413.7)	190.5 (162.6 to 225.6)	−0.9 (−0.94 to −0.86)
Andean Latin America	34 609.9 (28 878.7 to 42 124.1)	193.9 (161.8 to 237.3)	77 363 (64 579.9 to 93 681.4)	137.3 (114.8 to 166)	−1.17 (−1.2 to −1.15)
Australasia	13 344.7 (11 385.3 to 15 772.9)	58.4 (50 to 68.4)	30 385.9 (25 548.4 to 35 983.8)	50.1 (42.1 to 59.3)	−0.4 (−0.45 to −0.34)
Caribbean	42 425.9 (35 190.3 to 51 049)	174.2 (146 to 209.1)	70 581.5 (59 152.8 to 84 587.1)	130 (109 to 156.1)	−0.93 (−0.95 to −0.91)
Central Asia	55 141.1 (46 280.1 to 66 828.9)	132 (111.1 to 160.2)	72 637.6 (60 052.7 to 88 160.7)	105.9 (88.9 to 126.1)	−0.79 (−0.86 to −0.73)
Central Europe	67 897.4 (57 304.1 to 80 428.6)	49.2 (41.9 to 58.2)	92 636.6 (78 598.8 to 110 131.4)	38.6 (32.9 to 45.7)	−0.83 (−0.87 to −0.8)
Central Latin America	118 504.3 (100 711.8 to 139 490.1)	168.5 (143.9 to 200)	299 310.5 (255 750.4 to 353 856)	126 (107.4 to 148.9)	−0.91 (−0.96 to −0.87)
Central Sub-Saharan Africa	25 439.1 (21 147.5 to 30 717.6)	174.8 (146.9 to 209.4)	57 793.1 (48 292.6 to 68 716.2)	161.1 (135.6 to 191.4)	−0.2 (−0.27 to −0.13)
East Asia	602 900.5 (503 678.8 to 711 879)	88.3 (75.1 to 105.1)	1 137 135.4 (961 986.8 to 1 349 351.3)	53.7 (45.7 to 63.4)	−0.64 (−0.92 to −0.37)
Eastern Europe	226 898 (195 997.2 to 266 370.9)	86.9 (75.7 to 102.6)	252 879.1 (218 326.1 to 297 491.6)	69.3 (59.9 to 81.4)	−0.84 (−0.91 to −0.76)
Eastern Sub-Saharan Africa	166 898.4 (140 868.8 to 194 357.1)	293 (250.9 to 342.3)	308 159.2 (260 514.3 to 363 309.1)	236.7 (201.9 to 279.6)	−0.7 (−0.77 to −0.63)
High-income Asia Pacific	124 151.5 (108 382 to 143 458.7)	67.2 (59.1 to 77.3)	323 089.9 (283 086.4 to 376 459.5)	54.8 (47.8 to 62.8)	−0.61 (−0.71 to −0.51)
High-income North America	172 910.8 (150 019.5 to 201 313.8)	46.5 (40.4 to 53.9)	317 228.1 (274 947 to 368 049.7)	43.6 (37.7 to 50.6)	−0.22 (−0.36 to −0.07)
North Africa and Middle East	402 591.3 (339 256.2 to 477 781.2)	299.1 (254.9 to 357.8)	776 896.8 (655 149.2 to 918 338)	208.2 (176.6 to 247.2)	−1.22 (−1.26 to −1.18)
Oceania	2371.5 (1901.1 to 2901.9)	112.4 (92.4 to 136.1)	5139 (4231.3 to 6183.6)	92.2 (76.3 to 110)	−0.59 (−0.61 to −0.56)
South Asia	803 203.7 (669 772.6 to 948 045.4)	175.5 (148.5 to 207.6)	1 621 629.5 (1 354 035.6 to 1 917 154.9)	123.2 (104.4 to 145.6)	−1.44 (−1.56 to −1.33)
Southeast Asia	211 833.9 (176 766.2 to 251 299.4)	104.5 (87.7 to 124.4)	422 596 (354 981.9 to 499 060.8)	74.6 (62.8 to 88)	−1.15 (−1.19 to −1.12)
Southern Latin America	34 689.3 (29 150 to 41 989.7)	81.1 (68.5 to 98.9)	58 174.5 (48 817.7 to 68 864)	63.5 (53.3 to 75)	−0.73 (−0.76 to −0.7)
Southern Sub-Saharan Africa	53 767.6 (45 628.6 to 64 047)	236.3 (202.1 to 282.8)	96 003.9 (81 545.9 to 113 008.8)	202.2 (171.9 to 240.6)	−0.91 (−1.07 to −0.75)
Tropical Latin America	163 390.8 (140 039.4 to 189 585.9)	215.7 (185.1 to 252.1)	420 428.9 (360 786.5 to 492 901)	169.8 (145.3 to 199.3)	−0.3 (−0.46 to −0.13)
Western Europe	493 873.2 (416 915.4 to 583 523.7)	81.5 (69.2 to 95.7)	695 423.7 (592 196.5 to 812 825.1)	61 (51.9 to 70.8)	−0.92 (−0.99 to −0.85)
Western Sub-Saharan Africa	255 263.4 (216 161.3 to 295 046.2)	358.1 (307 to 418.7)	452 180.6 (383 396.6 to 527 663.6)	294.4 (250.8 to 345.3)	−0.71 (−0.77 to −0.65)

ASPR, Age-standardized prevalence rate; EAPC, Estimated annual percentage change; SDI, sociodemographic index; UI, uncertainty interval.

Gender disparities persisted, with males consistently showing higher prevalence and DALYs. Joinpoint regression revealed temporal fluctuations, indicating the steepest declines in ASPR during 1990–1994 and 2019–2021, while the most pronounced reductions in ASDR occurred between 2006 and 2010 (online supplemental figure 1A,B). Despite increasing absolute case numbers, both ASRs showed progressive declines throughout the study period (online supplemental figure 2A,B).

The glaucoma burden change during the COVID-19 pandemic (2019–2021)

Online supplemental figure 2 illustrates that the steepest increases in global glaucoma prevalence and DALYs occurred during 2010–2019, followed by a considerable slowdown between 2019 and 2021. This deceleration coincided with the most pronounced declines in the ASPR during the pandemic period, as identified by Joinpoint regression (online supplemental figure 1A). Regional analysis of three timeframes (1990–2021, 2010–2019 and 2019–2021) demonstrated minimal variation in the most recent period (2019–2021), with no significant increases in disease burden observed. Notably, high-middle SDI regions and East Asia exhibited absolute declines in both prevalence and DALYs during this period (online supplemental table 2).

Global distribution of disease burden due to glaucoma in 2021

In 2021, the global burden of visual impairment due to glaucoma exhibited substantial disparities. India reported the highest number of prevalent cases (1 372 410.29; 95% UI 1 145 859.64 to 1 620 636.16), followed by China (1 119 892.27; 95% UI 946 317.67 to 1 326 306.19), whereas Tokelau and Niue reported the lowest (1.30 and 1.59 cases, respectively) (online supplemental table 3). A similar trend was observed in DALYs, with India and China leading and Tokelau and Niue showing minimal impact, reflecting demographic scaling due to population size differences (figure 1A).

Figure 1. Geospatial mapping of glaucoma-induced visual impairment in 204 countries and territories (2021). Prevalence number and DALYs number (A), ASPR and ASDR per 100 000 populations (B), EAPC in ASPR and EAPC in ASDR (C). ASPR, age-standardised prevalence rates; ASDR, age-standardised DALY rates; DALYs, disability-adjusted life years; EAPC, Estimated annual percentage change.

The age-standardised glaucoma burden was most severe in African nations, particularly Niger (ASPR 345.65 per 100 000, ASDR 41.83 per 100 000) and Nigeria (ASPR 320.85 per 100 000), and lowest in certain European and Asian regions, notably Taiwan province of China (ASPR 17.63 per 100 000, ASDR 0.59 per 100 000) and Sweden (ASPR 23.46 per 100 000). Botswana also reported a high ASDR of 38.31 per 100 000, whereas North Korea showed a low ASDR of 1.58 per 100 000 (online supplemental table 3, figure 1B).

Over 30 years, most countries experienced declines in ASPR and ASDR. However, The Gambia, Benin, Niger and others were among those with significant increasing trends, whereas Tunisia and the Maldives demonstrated the most substantial reductions (figure 1C).

Age-specific burden of glaucoma

Figure 2 depicts the global age-specific prevalence of glaucoma-related visual impairment in 2021, characterised by a complete absence of reported cases below the age of 45 years. Both prevalence and DALYs exhibited a characteristic rise-and-fall pattern, peaking in the 70–74 age group (figure 2A and C). In contrast, the corresponding crude rates demonstrated a steady, monotonic increase with advancing age (figure 2B and D). A clear gender disparity was observed: males exhibited higher prevalence until age 80, after which the burden shifted to females, a pattern consistent with established differences in life expectancy between the sexes.

Figure 2. Age-sex stratified global burden of glaucoma-induced visual impairment (2021): prevalent cases (A) and prevalence rate trends (B). DALYs counts (C) and DALYs rate trends (D). e+05 represents×105. DALYs, disability-adjusted life years.

Over 30 years, global glaucoma prevalence and DALYs increased substantially. Decomposition analysis identified population growth as the primary driver of rising prevalence, accounting for 533.91% of the net increase (online supplemental figure 3A). In contrast, the increase in DALYs was principally driven by population ageing (102.59%), supplemented by population growth (83.07%); these upward pressures were partially counterbalanced by favourable epidemiological changes, which exerted an offsetting effect of −85.66% (online supplemental figure 3B). Collectively, these findings underscore the profoundly age-dependent nature of glaucoma and demonstrate how the compounding effects of demographic ageing and population expansion have shaped the escalating disease burden. These overarching trends were consistently observed in both sexes (online supplemental figure 3).

Disease burden of glaucoma stratified by SDI

A consistent inverse association was observed between the ASDR and the SDI across all 21 GBD regions and globally (figure 3A). Regions with higher SDI values generally exhibited lower ASDR, whereas lower SDI regions demonstrated disproportionately higher rates. This pattern was particularly pronounced in Western Sub-Saharan Africa, North Africa and the Middle East and Southern Sub-Saharan Africa, where the observed ASDR substantially exceeded expected values, a disparity potentially attributable to socioeconomic constraints. At the national level, ASDR typically declined with increasing SDI; however, several notable outliers—namely Botswana, Iran, Libya, Saudi Arabia and Qatar—maintained disproportionately high ASDR despite their higher SDI rankings (figure 3B). These findings highlight persistent and substantial disparities in the global burden of glaucoma.

Figure 3. Glaucoma-related visual impairment burden stratified by SDI: 21 Global Burden of Disease regions (A), 204 countries and territories (B). SDI: sociodemographic index. DALYs, disability-adjusted life years.

Disease burden due to glaucoma predicted by ARIMA model

The ARIMA model quantified projected trends in glaucoma prevalence and DALYs from 2022 to 2050. Residual analysis confirmed model adequacy, supported by a Ljung–Box test result greater than 0.05. A mean absolute percentage error below 10% indicated excellent predictive performance (online supplemental table 4). Global projections suggest glaucoma prevalent cases may rise to 9 423 763.56 (95% UI 3 237 172.34 to 15 610 354.78) and DALYs to 1 002 120.46 (95% UI 923 581.73 to 1 080 659.2) by 2050 (online supplemental table 5, figure 4A,B). Conversely, ASR shows progressive declines: ASPR to 63.41 (95% UI 48.55 to 78.26) and ASDR to 5.44 (95% UI 1.76 to 9.12) per 100 000 (online supplemental table 5, figure 4C,D). These consistent trends across sexes (onlinesupplemental figures 4 5) reflect a persistent yet gradually improving profile of the global glaucoma burden.

Figure 4. Autoregressive Integrated Moving Average model projections of glaucoma-attributable visual impairment burden (2022–2050). Prevalence number of both (A), DALYs number of both (B), ASPR of both (C), ASDR of both (D). e+05 represents×105, e+06 represents×106. Yellow dots and shadow areas indicate forecast trends and 95% CI. ASPR, age-standardised prevalence rates; ASDR, age-standardised DALY rates; DALYs, disability-adjusted life years.

Discussion

This study provides a systematic global assessment of glaucoma-related vision loss from 1990 to 2021, analysing trends by age, sex and socioeconomic development across 204 countries and 21 GBD regions. Employing ARIMA modelling, we projected glaucoma burden through 2050. Both prevalence and DALYs rose steadily, with the sharpest increases occurring between 2010 and 2019. Decomposition analysis indicated that population growth drove rising prevalence, whereas population ageing and demographic expansion principally accounted for the increase in DALYs. These trends may also reflect better case detection due to advances in glaucoma screening. Meanwhile, ASPR and ASDR declined consistently over the study period.

Our study revealed distinct epidemiological trends, characterised by declining rates in most regions, contrasted with persistent or increasing burdens in several African nations, including The Gambia, Benin and Niger. This pattern is consistent with growing evidence of a disproportionately high and increasing glaucoma burden in Africa.^{13 20} Meta-analyses confirm highest prevalence in Black populations,^{4 21} with ultraviolet radiation exposure potentially exacerbating the disproportionate burden in Africa.²⁰ These findings warrant prioritised international resource allocation to high-burden African regions. The persistence of significant global disparities in glaucoma-related vision loss, despite decades of policy implementation and healthcare improvements, highlights the imperative for targeted interventions that address both biological determinants and structural inequities in healthcare access.

Glaucoma is an age-related blinding eye disease that has emerged as one of the leading causes of visual impairment and blindness, particularly in the context of global population ageing.^{4 22} Our study demonstrates a progressive increase in both prevalence and DALYs with advancing age, peaking in the 70–74 years age group, which accounted for 17.13% of global glaucoma cases and 16.70% of DALYs, consistent with the findings of Lin et al.¹² Notably, this trend reverses beyond 74 years of age, likely attributable to declining population size in older age strata, particularly given the current global average life expectancy of 73.3 years (WHO, 2023). However, both crude prevalence rates and crude DALYs demonstrate persistent upward trends, suggesting a potentially increasing disease burden in ageing populations. Mechanistic studies suggest that age-related fibronectin accumulation in the trabecular meshwork may impair aqueous humour outflow, contributing to elevated IOP and glaucoma susceptibility in the elderly.²³ Juvenile open-angle glaucoma, a rare subtype of primary glaucoma, typically manifests before 40 years of age and imposes significant morbidity on younger individuals.²⁴ Intriguingly, the GBD 2021 study lacks burden estimates for individuals under 45 years of age, precluding a formal assessment in this demographic. Nevertheless, early screening of high-risk populations and prompt therapeutic intervention remains critical to mitigating disease progression.

Existing evidence demonstrates significant sex differences in glaucoma burden. Previous meta-analyses have reported that males exhibit a 1.36 to 1.37-fold higher prevalence of primary open-angle glaucoma (POAG) than females.^{4 25} By contrast, our analysis of the GBD 2021 database, which encompasses all glaucoma subtypes, revealed a more modest male-to-female prevalence ratio of 1.02. This discrepancy probably arises from the inclusion of heterogeneous glaucoma types in the GBD database, as opposed to the restricted focus on POAG in previous syntheses. Our findings indicate that male glaucoma case numbers exceed female figures before age 80, while DALYs remain elevated in males until age 85, after which this pattern reverses. This epidemiological transition may be explained by the well-documented shorter life expectancy of males relative to females. Nevertheless, females consistently demonstrated lower values across all evaluated metrics, including crude prevalence, absolute DALYs, ASPR and ASDR, while maintaining parallel temporal trends with males. These observations align with previous GBD studies.^{11 13 26} From a public health perspective, these results underscore the need for sex-specific intervention strategies. Policymakers should consider prioritising the following measures: first, implementing comprehensive public education initiatives to enhance glaucoma awareness; second, expanding screening programmes that target high-risk populations, particularly males in their eighth decade of life and individuals with a family history of glaucoma; and third, developing evidence-based clinical guidelines to optimise early detection and resource allocation. Such targeted approaches could substantially mitigate the disproportionate burden observed in specific demographic subgroups.

Significant geographic disparities in glaucoma burden were observed across the SDI levels. Western Sub-Saharan Africa, a low-SDI region, demonstrated persistently high ASDR, particularly in Niger (ASDR: 41.83 per 100 000; SDI: 0.17). Notably, Botswana, Iran and Libya exhibited elevated ASDR despite higher SDI rankings. This pattern may be attributed to a combination of factors such as accelerated population ageing, ethnic-specific genetic susceptibility, inadequate screening coverage and, in some cases, healthcare system disruptions associated with regional instability.

Consistent epidemiological patterns emerge across multiple studies examining glaucoma burden disparities. The World Bank’s burden of disease analysis reported substantially higher glaucoma-related ASDR in low- and middle-income countries than in high-income nations.²⁷ Similarly, a study by Freeman et al encompassing 70 countries and territories found the prevalence of visual impairment from all causes to be three times greater in low-income settings.²⁸ Both sets of findings align closely with our results. The observed gradient in disease burden is primarily attributable to structural determinants, such as limited healthcare access, particularly in rural areas, lower educational attainment and persistent poverty.^{29 30} However, the SDI alone constitutes an insufficient tool for comprehensive assessment, as demonstrated by the outlier nations. We therefore advocate for the adoption of multidimensional frameworks that integrate local epidemiological data, healthcare quality metrics, cultural perceptions of ocular health and geospatial service mapping. Such comprehensive approaches will facilitate the design of targeted interventions that address region-specific barriers to glaucoma care, thereby advancing beyond income-based classifications toward the implementation of precision public health strategies.

We observed a notable deceleration in the growth of glaucoma prevalence and DALYs during the 2019–2021 period, marked by the most pronounced decline in APC of ASPR, which reached −2.32. Notably, both high-middle SDI regions and East Asia demonstrated absolute reductions in glaucoma prevalence and DALY counts. This period coincided with the global COVID-19 pandemic, which was associated with significant excess mortality among elderly populations^{31 32} and the temporary suspension of non-emergency healthcare services.³³ As China represents a prototypical high-middle SDI country within East Asia, its stringent pandemic containment measures may have delayed glaucoma case presentations.³⁴ Several clinical studies have confirmed a decrease in the number of glaucoma patients attending ophthalmology visits during the COVID-19 pandemic.^{35 36} Although these factors may have introduced systematic downward bias in burden estimates, the precise drivers of the observed deceleration require further investigation. These findings highlight the need for vigilance regarding potential post-pandemic diagnostic catch-up effects that could transiently elevate recorded incidence rates. They also underscore the importance of implementing remote IOP monitoring and integrating chronic eye disease management into public health emergency preparedness plans.

Glaucoma remains the second most common cause of global blindness and the fourth leading cause of moderate-to-severe vision impairment.³⁷ Our ARIMA modelling predicts a 24.2% increase in global prevalence and a 31.88% rise in DALYs by 2050 compared with 2021 levels, although these growth rates represent a considerable slowdown relative to previous decades. Projected declines in ASR (ASPR: −29.62%; ASDR: −40.21%) suggest potential benefits from improved diagnostics, healthcare access and socioeconomic development. However, these projections require cautious interpretation given countervailing demographic trends of population ageing and increased longevity that may inflate future burden estimates.³⁸ These nuanced forecasts provide policymakers with an evidence-based framework for dynamic resource allocation, supporting the development of adaptive glaucoma management programmes that can respond to evolving population health dynamics.

Several limitations merit consideration. First, the GBD 2021 data’s exclusion of under-45 populations limits early-onset disease analysis, while aggregated estimates without subtype stratification constrain clinical applications. Second, the global scope of the GBD database encompasses 204 countries and territories, yet it faces challenges regarding inconsistent data quality and completeness. In regions characterised by a low SDI, for example, Western Sub-Saharan Africa, a lack of local epidemiological surveys leads to a greater reliance on modelled estimates. This reliance may introduce residual bias when compared with regions with high SDI levels, which typically benefit from more robust primary data sources. An additional limitation stems from the DisMod-MR 2.1 modelling framework, whose outputs depend directly on the quantity and quality of available scientific literature. As a result, estimates for regions with few published studies on glaucoma are associated with wider uncertainty intervals for both prevalence and DALYs. Third, pandemic-related care disruptions may have caused prevalence underestimation, as non-emergency ophthalmic visits declined globally, and diagnostic variability across healthcare systems could affect the accuracy of DALYs calculations. Fourth, while robust for trend analysis, our models cannot fully adjust for unmeasured confounders such as regional treatment disparities, nor can they simultaneously evaluate key demographic and socioeconomic drivers. They also lack the capacity to rigorously quantify the impact of discrete events such as the COVID-19 pandemic. Future studies should leverage advanced frameworks like mixed-effects models, interrupted time series or machine learning to better account for confounding, model complex trends and evaluate interventions. Despite these constraints, our findings provide policymakers with evidence for targeted resource allocation, particularly in high-burden regions. Subsequent iterations of the GBD study that integrate refined diagnostic criteria and harmonised reporting standards hold promise for enabling more precise and comparable burden estimates in the future.

Conclusion

Analysis of the 2021 GBD database reveals that the global glaucoma burden has declined over the past three decades, although unevenly across regions. Areas with a lower SDI sustained a higher disease burden. Males exhibit greater susceptibility to glaucoma; however, female prevalence surpasses that of males after the age of 80, with the risk increasing significantly from 60 years onward. Data from the COVID-19 pandemic period requires cautious interpretation, and post-pandemic surges in glaucoma cases warrant continuous monitoring. Projections to 2050 suggest persistently high prevalence and DALYs, mainly driven by population ageing despite advances in diagnosis, which highlights the need for effective strategies to reduce incidence. These findings emphasise the importance of enhancing early detection, diagnosis and treatment in high-burden regions to reduce future impact.

Data availability statement

Data are available in a public, open access repository. Data are available upon reasonable request.