INTRODUCTION
Cataracts constitute a major public health challenge as the leading cause of global blindness. Data from the Global Burden of Disease (GBD) study show that cataracts persistently rank as the primary cause of blindness worldwide, accounting for approximately 15.2 million cases and representing 45% of the global blindness burden1. Given that cataract treatment is highly dependent on surgical intervention, it imposes a heavy socioeconomic burden, particularly in regions with scarce healthcare resources2. Previous research has identified numerous likelihood factors for cataract, including age, sex, ethnicity, and environmental factors such as ultraviolet radiation exposure and tobacco smoke exposure3. Therefore, identifying and proactively controlling modifiable likelihood factors for cataract is imperative.
Smoking represents one of the major global public health threats, causing over 7.5 million deaths annually worldwide4. Substantial evidence indicates that tobacco smoke exposure is strongly associated with several blinding eye diseases, including cataract, glaucoma, and age-related macular degeneration5,6. Experimental studies have revealed potential mechanisms involving tobacco-induced oxidative stress and subsequent photochemical reactions, leading to lens injury and accelerated cross-linking denaturation of α-crystallin proteins7. Epidemiological studies have consistently confirmed the association between smoking and an elevated likelihood of cataract8-11. Compared to never smokers, individuals with a history of smoking exhibit a significantly higher prevalence ratio for cataract and a higher rate of undergoing cataract surgery12. Furthermore, a large prospective cohort study demonstrated that exposure to tobacco smoke increases the risk of developing cataract; compared to never smokers, individuals exposed during childhood, adolescence, and adulthood exhibited increased risks9.
However, existing studies exhibit significant limitations: they lack long-term spatiotemporal dynamic analyses of the years lived with disability (YLDs) burden attributable to smoking-related cataract stratified by the sociodemographic index (SDI); the impact of tobacco control policies on the disease burden has not been systematically quantified, hindering the evaluation of public health intervention effectiveness; and current predictive models fail to integrate the complex interaction effects of population aging and declining smoking rates, limiting accurate projections of future burden.
Leveraging GBD 2021 data, this study quantifies the visual impairment burden from smoking-related cataracts (1990–2021) across geography, SDI, age, and sex, with projections to 2050, to provide deeper insights into the global disease burden.
METHODS
Data source
This study is a secondary analysis based on the GBD 2021 database. Data were sourced from the publicly accessible database hosted by the Institute for Health Metrics and Evaluation (IHME) at the University of Washington (https://ghdx.healthdata.org/gbd-2021). This database has undergone institutional ethics review, and prior studies have confirmed it meets criteria for exemption from ethics review13. The GBD 2021 study integrates multiple heterogeneous data sources – including epidemiological surveys, disease registries, electronic health records, and scientific literature – and employs Bayesian meta-regression tools, particularly DisMod-MR 2.16, to impute and smooth estimates across locations and over time. This approach robustly handles missing and sparse data, producing complete and consistent estimates of mortality, morbidity, and health loss globally.
Definitions and stratification
Smokers were defined as current or former daily or non-daily users of any smoked tobacco product (including cigarettes, kreteks, pipes, waterpipe, cigars, bidis, and other traditional smoked products). Users of smokeless tobacco, e-cigarettes, and heated tobacco products were excluded. Cataract was defined according to the International Classification of Diseases, Tenth Revision (ICD-10) codes (H25-H26.9, H28-H28.8) as a pathological state characterized by abnormal protein aggregation leading to lens opacity and visual impairment14. Countries and territories were stratified using the SDI15. The SDI is a composite measure of lag-distributed income per capita, average education level in the population aged ≥15 years, and the total fertility rate under the age of 25 years. Based on the SDI, 204 countries and territories were categorized into five quintiles: low, low-middle, middle, high-middle, and high SDI.
Data extraction and metrics
From the GBD 2021 database, we extracted the following metrics related to the burden of smoking-related cataract for the period 1990–2021, at the global level, by SDI quintile, and for 204 countries and territories: 1) Years lived with disability (YLDs), quantifying the burden of disability caused by the disease; and 2) Age-standardized YLDs rate (ASYLDsR), expressed per 100000 population, to eliminate the influence of population age structure. Data were inclusive of all sexes and age groups from 30 to ≥95 years. No additional exclusion criteria were applied. Age and sex were considered as potential confounders in the modeling process through stratification and standardization.
Statistical analysis
To quantify the spatiotemporal trends in the global burden of smoking-related cataract from 1990 to 2021, we used the ASYLDsR as the core metric and assessed its temporal dynamics using the EAPC. The EAPC was calculated based on a linear regression model applied to the natural logarithm of the ASYLDsR against calendar year: y = α + βx + ε, where y = ln (ASYLDsR), x is the calendar year, α is the intercept, β is the slope, and ε is the error term. The regression coefficient β was transformed using the formula: EAPC = 100 × [exp(β) – 1]. The direction of the trend was determined based on the value of β and its 95% confidence interval (CI): If β >0 and the lower limit of the 95% CI >0, the burden was considered to have an increasing trend. If β <0 and the upper limit of the 95% CI <0, the burden was considered to have a decreasing trend. If the 95% CI included 0, the trend was considered stable16. To visually present the temporal evolution of the disease burden, comparative analyses were performed using data from 1990 and 2021. Descriptive subgroup analyses were conducted to examine burden distribution patterns across sex, age groups (30–34 to ≥95 years), geographical regions, and SDI quintiles. The linearity assumption of the EAPC model was verified through visual inspection of log-linear plots and residual analysis. The decomposition analysis employed the classical Das Gupta method17 to attribute the change in YLDs from 1990 to 2021 to three factors: change in population size (solely due to change in number of population), change in population age structure (solely due to aging), and the relative contribution of epidemiological change to vision impairment-related YLDs caused by smoking-associated cataract. The projection of the future disease burden utilized the BAPC model. This model integrates age, period, and birth cohort effects to analyze the complex interactions of demographic factors on disease burden events. Its Bayesian framework effectively handles complex data structures and uncertainty18. All analyses were conducted in R Studio (version 4.3.3). Key metrics are reported with their corresponding 95% uncertainty intervals (UI). UIs are used throughout the GBD study to reflect stochastic uncertainty from multiple modeling steps, while CIs are derived from regression models. UIs are presented for burden estimates, and CIs for trend statistics.
RESULTS
Global burden trends
In 2021, the global YLDs attributable to smoking-related cataract reached 225174 cases (95% UI: 151200–325529), representing a 38.9% increase compared with 1990 (162083 cases; 95% UI: 108854–230787). Concurrently, the ASYLDsR significantly declined, decreasing from 8.62 per 100000 population (95% UI: 5.81–12.26) in 1990 to 4.11 per 100000 (95% UI: 2.94–5.63) in 2021, with an EAPC of -1.41% (95% CI: -1.48 – -1.33) (Table 1 and Figure 1). Decomposition analysis revealed significant disparities in the drivers of YLDs growth: population growth (+202.47%) and population aging (+46.31%) acted as positive drivers, while epidemiological changes (-148.78%) exerted a negative effect (Figure 2, and Supplementary file Table 1). Population growth contributed most substantially in low-middle SDI (+364.30%) and low SDI (+363.47%) regions. Population aging had the most pronounced impact in high SDI regions (+165.55%), whereas the negative effect of epidemiological changes was most prominent in high SDI regions (-417.03%).
Table 1
Estimated YLDs cases, age-standardized YLDs rate per 100000 population, and temporal trends of vision impairment due to smoking-related cataract, by sex, age group, SDI, and region, 1990–2021
[i] YLDs: years lived with disability. UI: uncertainty intervals. CI: confidence interval. SDI: sociodemographic index. EAPC: estimated annual percentage change. Countries/regions by SDI quintile: Low (34), Low-middle (48), Middle (41), High-middle (48), High (33). Data source: Global Burden of Disease 2021.
Figure 1
Trends in age-standardized YLDs rates (A) and number of YLDs cases (B) for smoking-related cataract globally and by SDI quintile, 1990–2021. Trends were smoothed using LOESS regression for visual representation. Countries/regions by SDI quintile: Low (34), Low-middle (48), Middle (41), High-middle (48), High (33)
Figure 2
Changes in smoking-related cataract YLDs attributable to population-level determinants (population growth, aging, and epidemiological change) from 1990 to 2021 at the global level and by SDI quintile. The black dot represents the overall value of change contributed by all 3 components. Countries/regions by SDI quintile: Low (34), Low-middle (48), Middle (41), High-middle (48), High (33)
Sex and age stratification
The number of cases was consistently higher among males than females. Male YLDs increased from 131529 cases (95% UI: 88071–187121) in 1990 to 184884 cases (95% UI: 124009–266453) in 2021. The male ASYLDsR decreased from 7.38 per 100000 (95% UI: 5.3–10.1) to 4.59 per 100000 (95% UI: 3.24–6.39) over this period (EAPC= -1.46%; 95% CI: -1.52 – -1.40). Female YLDs increased from 30554 cases (95% UI: 19852–44402) in 1990 to 40290 cases (95% UI: 25965–60938) in 2021, with the ASYLDsR declining from 1.45 per 100000 (95% UI: 1.05–1.97) to 0.87 per 100000 (95% UI: 0.61–1.23) (EAPC= -1.60%, 95% CI: -1.77 – -1.44) (Table 1, and Supplementary file Figure 1). Age distribution analysis showed that both the number of cases and the age-standardized rate increased with age. The 65–69 years age group bore the heaviest burden in 2021 (36922 cases; 95% UI: 24621–53341 and ASYLDsR=13.39 per 100000; 95% UI: 8.93–19.34), exhibiting an annual rate reduction of -1.19% (95% CI: -1.32 – -1.06). While the 30–34 years age group had the largest absolute number of cases, it also showed the steepest decline in ASYLDsR, falling from 0.74 per 100000 (95% UI: 0.43–1.08) in 1990 to 0.33 per 100000 (95% UI: 0.19–0.49) in 2021 (EAPC= -2.84%; 95% CI: -3.03 – -2.64) (Table 1 and Figure 3).
Figure 3
Number of YLDs cases (A) and age-standardized YLDs rates (B) for smoking-related cataract by age group and sex, 2021. Trends were smoothed using LOESS regression for visual representation. The shaded areas represent the 95% uncertainty intervals. Trends were smoothed using LOESS regression for visual representation.
SDI and geographical heterogeneity
High SDI regions had the lowest ASYLDsR in 2021 (0.89 per 100000; 95% UI: 0.6–1.31), significantly lower than in 1990 (1.42 per 100000; 95% UI: 0.99–2.03) (EAPC= -1.60%; 95% CI: -1.70 – -1.50). Conversely, low-middle SDI regions exhibited the highest ASYLDsR in 2021 (5.22 per 100000; 95% UI: 3.72–7.18) but also the largest rate reduction (EAPC = -2.08%; 95% CI: -2.14 – -2.02) (Table 1 and Figure 1). SDI level was inversely associated with ASYLDsR (Supplementary file Figure 2). Country-level analysis revealed that high-income countries experienced concurrent decreases in both YLDs cases and ASYLDsR. In contrast, low- and middle-income countries saw an increase in YLDs cases alongside a decrease in ASYLDsR (Supplementary file Figures 3 and 4). The burden was particularly pronounced in Asian countries, with China (60057 cases) and India (59155 cases) collectively contributing 52.9% of the global YLDs total in 2021 (Supplementary file Table 1).
DISCUSSION
Based on the GBD 2021 study, this research systematically evaluated the spatiotemporal evolution of the global burden of vision impairment due to smoking-related cataract from 1990 to 2021. The results indicate that although the global ASYLDsR for smoking-related cataract showed a significant declining trend, the number of YLDs cases increased. This phenomenon is primarily driven by population growth and aging. However, epidemiological changes related to tobacco exposure – including declining smoking prevalence, increased smoking cessation rates, and improved access to cataract surgery – have partially offset this increase4,19. This finding underscores the profound impact of global demographic transition on disease burden, demonstrating that even with improvements in age-standardized rates, the absolute disease burden may still rise due to population expansion and intensifying aging15.
Demographic decomposition analysis further revealed regional heterogeneity. Population growth contributed a substantial 364% to the increase in YLDs in low-middle SDI regions, reflecting their early-stage demographic transition characteristics. In contrast, the significant contribution of population aging and the strong negative effect of epidemiological changes in high SDI regions confirm the transformative role of tobacco control legislation, universal health coverage, and high-quality ophthalmic services. These factors have successfully shifted smoking-related cataract from a highly disabling condition towards a manageable disease4. Furthermore, low-SDI countries, currently characterized by higher fertility rates and younger population structures, are transitioning from high fertility toward rapid aging20. Within this context, and given persistent inadequacies in healthcare resources21, the synergistic progress in tobacco exposure management and vision health services calls for sustained attention.
Sex and age disparities highlight heterogeneity in disease distribution. The burden of YLDs remained consistently higher among males than females; in 2021, the male ASYLDsR was 5.3 times higher than the female rate, and males accounted for 82% of global cases. This disparity closely associates with the global epidemiological pattern of significantly higher smoking prevalence among males compared to females4. A meta-analysis by Beltrán-Zambrano et al.22 confirmed that current smokers have a significantly increased risk of cataract, with the strongest association observed for nuclear cataract. Biologically, oxidative stressors in tobacco smoke, such as cadmium ions, accelerate the cross-linking denaturation of α-crystallin proteins in the lens23. This process may be exacerbated in males due to age-related declines in the lens’s antioxidant capacity23. It is noteworthy that the age-standardized rates declined continuously for both sexes, reflecting the combined success of global tobacco control policies and ophthalmic interventions24.
Age-stratified analysis revealed that the YLDs burden increased with age, peaking in the 65–69 years group, while the 30–34 years group exhibited the steepest decline in ASYLDsR. This trend aligns with the age-related nature of cataract, where oxidative damage to lens proteins accumulates over time, a process accelerated by smoking. Younger populations, with shorter cumulative exposure to tobacco and greater responsiveness to cessation interventions, experienced more significant declines in age-standardized rates4. This suggests that early interventions targeting younger populations, such as smoking cessation education, could effectively delay or prevent the onset of smoking-related cataract25. However, increases in youth smoking prevalence in low- and middle-income countries, along with the growing use of novel tobacco products such as e-cigarettes, could contribute to an elevated likelihood of early-onset cataract20, suggesting a need for targeted prevention approaches.
Disparities across SDI regions reveal global inequities in eye health. The age-standardized YLDs rate in low SDI regions was 3.1 times higher than in high SDI regions, reflecting the structural roots of imbalanced healthcare resource allocation. High SDI countries, through comprehensive tobacco control policies (such as Australia’s plain packaging legislation and the UK’s universal smoking cessation services) and efficient cataract surgical networks, have achieved a ‘low-level stable’ status for smoking-related cataract26. Conversely, low SDI regions face severe challenges in their prevention and control systems due to scarce healthcare resources leading to low cataract surgical coverage, weak implementation of tobacco control policies24,27, and persistently high smoking prevalence4. Furthermore, the decline in age-standardized rates was faster in low SDI regions compared to high SDI regions. This phenomenon may stem from the greater marginal benefits of interventions in settings with a high baseline burden26. Decomposition analysis further corroborates the mechanisms of regional heterogeneity: population expansion was the primary driver of YLDs growth in low SDI regions, while aging dominated in high SDI regions. This disparity suggests that regionally tailored strategies may be needed: low-SDI regions may need to prioritize strengthening tobacco control legislation and expanding primary eye care network coverage, whereas high-SDI regions should focus on optimizing cataract screening pathways and early surgical intervention for their aging populations28.
Heterogeneity at the national level highlights the complex distribution of the disease burden. Asian countries bear a high absolute burden due to their large population bases, with China and India collectively contributing 53% of global YLDs. Although both countries exhibited declining ASYLDsR, accelerating population aging is driving continued growth in the absolute burden. China faces unique challenges: the smoking prevalence among individuals aged ≥40 years was 27.2% in 2019–2020, the male smoking rate was 21 times that of females, and per capita tobacco consumption exceeded the global average29; simultaneously, cataract surgical rates in rural areas remain significantly lower than those in high-income countries30. In contrast, high-income countries have achieved rapid reductions in disease burden through stringent tobacco control regulations (e.g. smoke-free public places, tobacco taxation) and highly efficient surgical service systems31,32.
Limitations
This study has several limitations. First, as the GBD study relies on modeled estimates, the quality and coverage of primary data in some countries – particularly in low SDI regions – may be limited, which could introduce estimation bias. Second, the calculation of YLDs is based on population attributable fractions (PAF), the accuracy of which depends on the quality of smoking exposure data and the relative risk estimates for the association between smoking and cataract. Although GBD uses standardized Bayesian models to quantify uncertainty, some residual uncertainty remains. Third, variations in cataract diagnostic criteria across regions and over time may affect consistency in estimates. While GBD attempts to harmonize these differences through disease modeling, cross-regional comparability may still be compromised. Fourth, despite rigorous adjustment in GBD methods, residual confounding may persist. Fifth, underreporting of smoking – especially among women and youth – may lead to underestimation of the burden. Sixth, emerging trends such as e-cigarette use were not included due to a lack of sufficient long-term data. Finally, the ‘epidemiological change’ component in the decomposition analysis represents a residual term derived indirectly, encompassing both residual confounding and other non-demographic factors such as changes in diagnostic coding practices; thus, this component requires cautious interpretation. Moreover, the findings of this study reflect population-level associations and should not be interpreted as establishing causality.
CONCLUSIONS
The global burden of vision impairment due to smoking-related cataract remains a significant public health concern. Despite the declining trend in the global ASYLDsR, the absolute number of YLDs cases has increased, underscoring the underlying drivers of population growth and aging. The disease burden continues to be disproportionately heavy in low- and middle-income countries.
