This study is derived from the GBD 2021 dataset, a globally recognized epidemiological research platform that integrates multidimensional data from 204 countries and territories covering 371 diseases and health issues, and 88 risk factors between 1990 and 2021 (11). The GBD research network standardizes the data through cleaning, correction, and modeling processes to ensure comparability and reliability. The definition of LC is based on the International Classification of Diseases (ICD) coding system, primarily including C32-C32.9 in ICD-10 and corresponding codes 161–161.9, V10.21 in ICD-9. The database, built with the collaborative efforts of over 11,500 researchers, offers a comprehensive assessment system. For research on LC attributable to occupational exposure, mortality data was extracted from the Global Health Data Exchange (GHDx¹) and analyzed by factors such as age, gender (male working-age population in 15 to 64 years), and geographic distribution. To further investigate the impact of socio-economic factors on deaths burden, the study incorporated the SDI framework². The SDI, which is based on per capita income, education level, and fertility rate, divides regions into five development tiers, ranging from low to high.

To better understand the distribution characteristics of deaths burden and its influencing factors, the research explores demographic and socio-economic factors, with a particular focus on the impact of population structure and the SDI on the variation in deaths burden across different regions. In terms of trend analysis, a longitudinal research design was employed, using long-term data from 1990 to 2021. The study evaluates temporal trends in deaths burden by calculating the estimated annual percentage change (EAPC). Which is calculated based on a logarithmic linear regression model, with the core formula being: ln(ASR) = a + bx + e, where ln(ASR) represents the natural logarithm of the ASR, x is the year variable, a is the model intercept, b is the regression coefficient, and e is the error term. The EAPC is calculated from the regression coefficient b using the formula: EAPC = (exp(b) – 1) × 100. To assess the statistical significance of trend changes, the 95% confidence interval (CI) was calculated. A CI upper bound below 0 indicates a significant decreasing trend in deaths burden, while a lower bound above 0 suggests a significant increasing trend (12).

This study employs a decomposition analysis method to investigate the driving factors in the deaths cases of LC attributable to occupational exposure. It systematically evaluates the impact mechanisms of population structure changes, population size expansion, and the evolution of epidemiological characteristics on deaths burden from 1990 to 2022. Following Das Gupta’s decomposition framework, changes in deaths (ΔD) were partitioned into three components: ΔD=D2021 − D1990 = Δage+Δpop+Δepi, where Δage (aging), Δpop (population growth), and Δepi (epidemiological rate changes) were calculated via counterfactual scenarios (13). Notably, changes in epidemiological characteristics, including improved access to healthcare and the widespread adoption of preventive measures, have had complex effects on deaths burden. Through quantitative analysis, this study identifies the contributions of each factor to the deaths cases of LC, providing a scientific basis for a comprehensive understanding of the epidemiological evolution.

This study employs the Bayesian Age-Period-Cohort (BAPC) model to predict the evolving trends in the deaths burden of LC attributable to occupational exposure. The prediction model, implemented on the R language platform using the Integrated Nested Laplace Approximation (INLA, 23.09.09) algorithm and the BAPC (0.0.36) package, is capable of effectively handling complex hierarchical data structures. The model not only captures the disease risk characteristics across different age groups but also reflects the influence of social environmental changes and medical advancements on disease patterns (14). Compared to traditional prediction methods, the BAPC model offers superior explanatory power and predictive accuracy, particularly in long-term trend analysis and intergenerational comparisons. This makes it a reliable analytical tool for gaining deeper insights into the epidemiological characteristics of LC attributable to occupational exposure. Using this modeling framework, we forecast mortality rates from 2022 to 2040, which will provide crucial evidence for optimizing health resource allocation and formulating disease prevention strategies.

This study employs an advanced analytical approach to construct an evaluation model based on ASDR, aimed at exploring the association between the deaths burden and the SDI. Compared to traditional linear regression models, the frontier analysis method offers significant advantages: it not only identifies multidimensional driving factors influencing deaths burden but also effectively captures the complex non-linear relationship between SDI and deaths burden. Methodologically, this study introduces an innovative approach by combining Locally Weighted Regression (LOESS) with Local Polynomial Regression, using a smoothing span parameter of 0.3 to construct a non-linear boundary curve between SDI and ASDR. This boundary curve holds substantial theoretical significance, as it establishes the achievable theoretical minimum ASDR for each country or region at its specific development level, providing a scientific benchmark for evaluating national disease prevention and control performance. Additionally, by quantifying the gap between the observed disease burden and the theoretical minimum value, this method precisely identifies areas that require focused improvement (15). In order to validate the robustness and reliability of the results, we performed 1,000 bootstrap resamples, calculating the average ASDR for each SDI value to effectively control for data fluctuations. In the evaluation phase, we constructed a quantifiable assessment indicator system by measuring the absolute distance between each country’s actual ASDR in 2021 and the boundary curve. This indicator not only reflects the current effectiveness of national disease control but also evaluates the potential for future reductions in the LC attributable to occupational exposure burden.

This study uses standardized epidemiological indicators to quantify the deaths burden, with ASDR calculated per 100,000 person-years and reported alongside their 95% Confidence Interval (CI) to reflect the precision of the predictions. To ensure the robustness of the findings, multiple statistical models and techniques were applied to analyze temporal trends, decompose contributing factors, project future patterns, and evaluate disparities across sociodemographic levels. Each model was selected based on its suitability for the corresponding research objective, and all outputs were visually represented through standardized plots and comparative graphs. Additionally, the EAPC is also provided with a 95% CI. The entire data analysis and visualization process was performed using R Studio (Version 4.3.3 for Windows). For statistical inference, we used two-tailed tests, with a p-value < 0.05 indicating significance.

In 2021, the global male LC attributable to occupational exposure to asbestos resulted in a total of 533.63 deaths (95% CI: 280.15, 842.41), representing a 22% decrease compared to 1990. The ASDR was 0.02 per 100,000, with an average annual decrease of −3.23% since 1990 (Table 1 and Figure 1). For LC attributable to occupational exposure to sulfuric acid, the total number of male deaths in 2021 was 2131.15 (95% CI: 899.19, 3825.72), showing a 16.19% increase compared to 1990. The ASDR was 0.09 per 100,000, with an average annual decrease of −1.90% (Table 2 and Figure 2). Across different SDI regions, the High-middle SDI region had the highest number of male deaths cases from LC attributable to occupational exposure to asbestos in 2021, with 147.25 deaths (95% CI: 76.86, 234.07), while the Low SDI region had the lowest with 26.24 deaths (95% CI: 8.72, 53.90). The highest ASDRs were observed in the High and High-middle SDI regions, while the lowest were in the other SDI regions. The highest annual increase in ASDR was observed in the Low-middle SDI region (0.27%), and the fastest decrease was in the High-middle SDI region (−4.49%) (Table 1 and Figure 1A). For male deaths from LC attributable to occupational exposure to sulfuric acid in 2021, the highest number was in the Middle SDI region (741.14), and the lowest was in the High SDI region (148.37). The highest ASDR was found in the Low-middle SDI region (0.15 per 100,000), and the lowest was in the High SDI region (0.03 per 100,000). The fastest annual decrease in ASDR was in the High-middle SDI region (−3.48%), while the slowest decrease occurred in the Low-middle SDI region (−0.68%) (Table 2 and Figure 2).

In 2021, an analysis of 21 GBD regions revealed significant regional differences in male LC attributable to occupational exposure. For occupational exposure to asbestos, the highest ASDR were reported in Southern Sub-Saharan Africa (0.12), Central Europe (0.07), and Tropical Latin America (0.07), while Oceania, High-income Asia Pacific, and Andean Latin America reported the lowest ASDRs (Table 1). From 1990 to 2021, Central Europe (0.29%) and Oceania (0.23%) exhibited an increasing trend in ASDR, while Australasia (−6.32%), High-income North America (−5.28%), and North Africa and the Middle East (−4.39%) experienced the most significant annual percentage decreases in ASDR (Table 1). Additionally, the comparison of differences in ASDR for LC attributable to occupational exposure to asbestos across different years in the 21 regions is shown in Figure 3A.

Similarly, in 2021, for male deaths from LC attributable to occupational exposure to sulfuric acid, the highest ASDRs were reported in the Caribbean (0.21), South Asia (0.18), and Tropical Latin America (0.18), while the lowest ASDRs were found in High-income Asia Pacific (0.01), Oceania (0.01), and Australasia (0.02) (Table 2). From 1990 to 2021, the Caribbean (0.82%) showed an increasing trend in ASDR, while High-income Asia Pacific (−5.19%), Western Europe (−4.22%), and Central Asia (−4.22%) exhibited the most significant decreases in annual percentage change (EAPC) (Table 2). The comparison of differences in male LC attributable to occupational exposure to sulfuric acid deaths across different years in the 21 regions is shown in Figure 3C.

At the regional level, we found a general positive correlation between SDI and the ASDR for LC attributable to occupational exposure to asbestos from 1990 to 2021. As SDI increased, the ASDR for LC showed an upward trend (Figure 3B). From 1990 to 2021, the ASDR in the Southern Sub-Saharan Africa region was significantly higher than its expected value, while the ASDR in Oceania was notably lower than expected during the same period (Figure 3B). A similar positive correlation was observed for LC attributable to occupational exposure to sulfuric acid (Figure 3D).

In 2021, the ASDR for LC attributable to occupational exposure to asbestos varied significantly across countries, ranging from 0 to 0.43 per 100,000 population. Lesotho reported the highest ASDR (0.43 per 100,000), followed by Monaco (0.20) and Eswatini (0.19) (Figure 4A). Between 1990 and 2021, the countries with the fastest average annual increases in ASDR were Saudi Arabia (27.47%), Georgia (12.98%), and El Salvador (9.57%). Conversely, the most significant average annual decreases were observed in Singapore (−8.09%), American Samoa (−7.81%), and the Republic of Korea (−6.97%) (Figure 4B).

For LC attributable to occupational exposure to sulfuric acid, the ASDR ranged from 0.01 to 0.38 per 100,000 in 2021. Cuba had the highest rate (0.38), followed by Pakistan (0.31) and Seychelles (0.25) (Figure 4C). From 1990 to 2021, the countries with the most rapid increases in ASDR included Solomon Islands (3.26%), Kiribati (2.82%), and Sri Lanka (2.64%), while the largest declines occurred in the Republic of Korea (−7.78%), Singapore (−6.10%), and Spain (−5.56%) (Figure 4D).

At the national level, in 2021, ASDR for LC attributable to asbestos exposure showed a positive correlation with SDI. Notably, Lesotho’s deaths burden was significantly higher than expected for its development level (Figure 5A). Similarly, for LC attributable to occupational exposure to sulfuric acid, Cuba’s burden was also markedly above the expected value based on its SDI (Figure 5B).

To explore the driving factors behind the changes in the deaths burden of LC attributable to occupational exposure, we assessed the effects of aging, population growth, and epidemiological changes over the past 32 years. Specifically, between 1990 and 2021, the global increase in LC attributable to occupational exposure to asbestos deaths was primarily driven by epidemiological changes (415.13%), while population growth (−216.16%) and aging (−98.96%) had offsetting effects (Figure 6A). In different SDI regions, aging was the largest driver of death cases in the middle SDI region (70.24%), while it had the greatest offsetting effect in the high-middle SDI region (−20.96%). Population growth was the primary driver of death cases in the middle SDI region (87.21%) and had the largest offsetting effect in the high-middle SDI region (−86.95%).

However, for LC attributable to occupational exposure to sulfuric acid, epidemiological changes played a protective role (−383.85%), while population growth (335.99%) and aging (147.86%) were the primary drivers of increased death cases (Figure 6B). In different SDI regions, aging was the largest driver of death case growth in the middle SDI region (92.78%), population growth was the main driver in the middle SDI region (125.47%), and epidemiological factors had the largest impact in the high SDI region (207.35%).

Figure 7 shows the projected number of deaths and ASDR for LC attributable to occupational exposure by 2040. Globally, it is expected that the number of deaths from LC attributable to occupational exposure to asbestos will decline annually (Figure 7A), reaching 372.78 by 2040. Similarly, the ASDR follows a downward trend, with the rate expected to decrease to 0.011 per 100,000 population by 2040 (Figure 7B). In contrast, for LC attributable to occupational exposure to sulfuric acid, the number of deaths is projected to increase, rising to 2,543.78 by 2040 (Figure 7C), while the ASDR will decrease to 0.078 per 100,000 population (Figure 7D).

Using data from 1990 to 2021, we conducted a frontier analysis related to SDI based on the ASDR to explore the potential for reducing the deaths burden of LC attributable to occupational exposure, considering the development levels of countries and regions (Figures 8A,C). The frontier curve represents the minimum (best-performing) ASDR observed for a given SDI, and the distance from the curve, i.e., the effective difference, quantifies the gap between the observed ASDR of a country or region and the achievable ASDR under optimal conditions. In 2021, the effective difference in ASDR for each country was calculated based on their SDI (Figure 8B). Overall, as SDI increases, the effective difference tends to be smaller, with variance stabilizing at higher SDI levels. Regarding occupational exposure to asbestos, in the 15 countries and regions with the greatest potential for improvement, the effective difference ranged from 0.037 to 0.179. These countries and regions include Lesotho, Monaco, Croatia, Eswatini, France, Namibia, Saint Kitts and Nevis, South Africa, Poland, Georgia, Slovenia, Montenegro, Turkey, Armenia, and Hungary (Figure 8B). For occupational exposure to sulfuric acid, the 15 countries and regions with the greatest potential for improvement include Cuba, Pakistan, Seychelles, Georgia, Romania, Saint Vincent and the Grenadines, Uruguay, Hungary, Zambia, Montenegro, Bulgaria, Bangladesh, Brazil, Haiti, and India (Figure 8D).

This study systematically reveals the global deaths burden of LC attributable to occupational exposure in the male working-age population from 1990 to 2021. The results show that globally, deaths from asbestos-related LC decreased by 22% (with an annual decline of 3.23% in ASDR), while sulfuric acid-related cases increased by 16.19% (with an annual decline of 1.90% in ASDR). High-SDI regions bore a heavier burden from asbestos exposure, while medium-SDI regions faced higher risks from sulfuric acid exposure. Decomposition analysis indicated that the increase in asbestos-related burden was primarily driven by epidemiological factors, while sulfuric acid-related risks were mainly influenced by population growth. By 2040, asbestos-related deaths are projected to decrease to 372 cases (ASDR 0.011/100,000), while sulfuric acid-related deaths are expected to rise to 2,543 cases (ASDR 0.078/100,000), indicating that medium-SDI regions will be a key focus for occupational health prevention and control in the male working-age population.

Previous study indicated that occupational exposure to carcinogens is a significant cause of death and disability worldwide, and asbestos is a major occupational risk factor. Over the past decades, the burden of occupational cancers has increased significantly, including LC (16). In the 2025 study on the disease burden of LC, three major attributable risk factors for LC were systematically reported, namely smoking, drinking alcohol, and occupational exposure (9). However, previous studies only explored the main attributable risk factors and did not conduct stratified research on the disease burden of each attributable risk factor. Based on previous studies, this research conducted an in-depth investigation into the global disease burden of the two major occupational risk factors for LC among the male labor force. The research on the attributed risk factors of different populations will be conducive to the precise prevention and control of LC. The results of this study show that deaths from asbestos related LC decreased, while sulfuric acid-related cases increased. Therefore, in the future, occupational protection against LC should pay more attention to occupational exposure to sulfuric acid. However, the confidence interval for ASDR is relatively wide. First, the precision of GBD modeling is inherently constrained by data scarcity in low-income regions; occupational exposure misclassification may further bias estimates. Second, while the decomposition model adjusts for smoking and alcohol use, residual confounding persists due to the lack of individual-level data on co-exposures. Third, the assumption of additive independence between demographic and epidemiological factors, though standard in decomposition frameworks, may oversimplify complex interactions in rapidly industrializing settings. These limitations necessitate cautious interpretation of ASDR estimates, particularly in regions with sparse mortality data.

Asbestos is widely used in construction and textiles for its strength, flexibility, heat resistance, and spinnability. Studies show that asbestos fibers, when deposited in the laryngeal mucosa via the respiratory tract, induce chronic inflammation, abnormal cell proliferation, genetic mutations, and, in combination with HPV, increase LC risk by 3.8 times (17). Since 2009, the IARC has classified asbestos as a Group 1 carcinogen for the larynx (18). Numerous studies have highlighted the link between asbestos exposure and LC. A 2020 occupational cancer study reported that asbestos was responsible for 63% of occupational cancer deaths from 2001 to 2016, with males accounting for 79% of these deaths (19, 20). Case–control studies in France and Germany have shown a dose–response relationship between asbestos exposure and LC (21, 22). Given asbestos’ long latency period (average 40 years) and peak use between 1930 and 1970, the peak of asbestos-related LC was expected between 1970 and 2010 (23). Our study shows a 22% decrease in asbestos-related LC deaths in the male working-age population globally from 1990 to 2021, reflecting global asbestos usage trends. The decline in asbestos-related LC deaths is largely attributable to global asbestos restriction policies, heightened occupational awareness, and improved protective measures. Since the International Agency for Research on Cancer (IARC) classified asbestos as a Group 1 carcinogen in 2009, over 60 countries have implemented full or partial bans on its use (18). High-SDI regions demonstrated the most significant ASDR reductions, reflecting stringent workplace regulations and asbestos substitution initiatives (19, 20). Conversely, middle/low-SDI regions with rising asbestos consumption still face increasing risks due to delayed policy implementation and inadequate protective measures. These findings underscore that legislative bans combined with occupational health interventions are effective in reducing asbestos-related cancer burdens, providing a critical roadmap for high-risk nations. Despite increasing recognition of its carcinogenicity, with organizations like ACGIH setting exposure limits, the ILO and WHO emphasize that no safe exposure level exists (24, 25). In recent years, India competes with China to be the largest consumer of asbestos in the world. Other countries among top user are Indonesia and the Russian Federation (10, 26). The import and recycling of asbestos is the chief source in India. In 2016–2017, India imported 310,592 tons of asbestos with 355,686 and 396,493 tons in previous years, respectively (26). Therefore, we call for a comprehensive ban on all types of asbestos worldwide and make the elimination of asbestos related diseases a top priority in the field of public health.

In this study, although high-SDI regions bear a heavier burden from asbestos exposure, decomposition analysis reveals that the primary driver behind this burden in these regions is epidemiological factors, such as improved disease detection. The decline in deaths burden since 1990 in high-SDI regions suggests that prevention and control measures have been effective. In 1983, Iceland became the first country to ban all forms of asbestos nationwide (27). Germany followed suit in 1993, and the use of asbestos in the region is now nearly zero, with over 50 other countries implementing similar bans (28). In recent years, the progress of national asbestos bans has slowed down, with some emerging countries lifting bans and others having long transition periods. This study shows that in low-SDI and some middle-SDI regions, the deaths burden decreased slowly from 1990 to 2021, with some middle-SDI regions even showing a slight increase. In sub-Saharan Southern Africa, the ASDR was significantly higher than expected, indicating insufficient prevention efforts. In middle- and low-SDI countries, the past decades of urbanization and industrialization likely led to extensive asbestos use, and these countries have focused primarily on acute and infectious diseases in their healthcare systems, neglecting sufficient attention to disease prevention and occupational protection.

Frontier analysis indicates that countries with the greatest improvement potential include Lesotho, Monaco, Croatia, Eswatini, France, Namibia, Saint Kitts and Nevis, South Africa, Poland, Georgia, Slovenia, Montenegro, Turkey, Armenia, and Hungary. To further reduce the deaths burden of asbestos-related LC in male working-age populations and eliminate asbestos-related diseases as a public health issue, greater efforts are needed in middle- and low-SDI regions. In countries like Lesotho, where the disease burden exceeds expectations for their development level, governments and public health personnel should conduct further epidemiological investigations to identify targeted prevention measures. Policy and prevention strategies can learn from successful cases in high-SDI regions, such as setting stricter exposure limits, increasing asbestos carcinogenicity education, and replacing asbestos with safer materials to achieve the goal of eliminating asbestos-related diseases at an earlier stage.

Sulfuric acid is widely used in industries such as chemicals, electroplating, and metallurgy, which predominantly involve male workers (7). Inorganic acid mist has been classified as a Group 1 carcinogen since 2012 (29). Although the carcinogenic mechanism of inorganic acid mist is not yet fully understood, low pH is a known pathogenic factor, causing genetic damage and inducing DNA damage in an acidic cellular microenvironment (30). Men, with higher smoking and alcohol consumption rates, may experience an additive effect from sulfuric acid exposure, increasing the risk of LC in the male working-age population. Additionally, higher miR-21 expression in the laryngeal epithelial cells of male patients enhances cell proliferation through the PTEN pathway, and poor adherence to occupational safety further exacerbates the risk (31). Previous studies have shown an increased incidence of LC among acid-pickers and workers in sulfuric acid petrochemical plants, with an exposure-response relationship (31). While cross-sectional and cohort studies at national and subnational levels reflect epidemiological patterns, they cannot fully capture the global and regional disease burden. Research on key global populations will aid in policy formulation and the implementation of prevention and control plans.

This study reveals that from 1990 to 2021, the burden of LC due to occupational sulfuric acid exposure was highest in low- and middle-SDI countries, while high-SDI regions saw a decline, likely due to stricter regulations and better workplace protections (32). Sulfuric acid mist, classified as a Group 1 carcinogen, poses risks primarily through inhalation, especially in poorly ventilated industrial settings. In many developing countries, rapid industrialization, aging populations, and inadequate occupational health systems have led to increasing disease burdens. Informal sectors often lack exposure controls, and regular health screening is limited (18). For example, in China, over 3 million workers in sulfuric acid-related industries have reached high-risk age groups, with aging accounting for over 90% of the disease burden. Countries like Cuba, Pakistan, and the Solomon Islands should strengthen prevention strategies, drawing on best practices from high-SDI nations (31). Targeted efforts are urgently needed in low- and middle-SDI regions to reduce long-term occupational exposure and improve early detection.

By 2040, deaths from occupational sulfuric acid-related LC are expected to rise, although ASDR will decline. For countries like Cuba, with a higher-than-expected burden, public health departments should conduct epidemiological studies to guide prevention measures. In the 15 countries with the greatest improvement potential, targeted prevention and equitable resource distribution are essential. We propose the following actionable measures: (1) low-cost exposure monitoring: Deploy affordable air quality sensors in LMIC workplaces to detect asbestos/sulfuric acid levels, with real-time alerts via SMS. (2) Tax incentives for safer alternatives: Offer 20–30% tax rebates to industries replacing asbestos with cellulose fibers or sulfuric acid with enzymatic cleaners. (3) Primary care task-shifting: Train community health workers to use smartphone-based laryngeal screening tools for early detection in high-risk groups. (4) Policy integration: Embed occupational exposure limits into existing national NCD prevention programs.

This study maps the global spatiotemporal characteristics of occupational sulfuric acid and asbestos-related LC in male working-age populations, filling a gap in occupational cancer epidemiology. By using decomposition models, it identifies 15 countries, such as Lesotho (asbestos) and Cuba (sulfuric acid), as key targets for intervention. However, limitations include: (1) data limitations due to underreporting in low-income countries; (2) historical bias as the study covers only 1990–2021 data, missing earlier high-exposure populations; (3) confounding factors, such as smoking and alcohol, not fully controlled for; (4) binary exposure classification, which cannot assess exposure gradients; (5) lack of exposure scenario differentiation, potentially missing key risk contexts. Future research should use multicenter cohort studies and dynamic policy models to improve burden assessments.