Global burden of smoking-attributable digestive cancers from 1990 to 2021: insights from the Global Burden of Disease study

Abstract

Background:

Smoking is causally linked to multiple cancers. We present global, regional, and national estimates of smoking-attributable digestive cancers (SADC) burden from 1990 to 2021 to inform smoking policy and cancer control.

Methods:

We analyzed data from the Global Burden of Disease (GBD) study, encompassing five subtypes of digestive cancers across 204 countries and territories. We estimated the mortality and disability-adjusted life years (DALYs) of SADC. Temporal trends were quantified by calculating the estimated annual percentage change (EAPC) of age-standardized rates.

Results:

Global SADC deaths increased from 347 533 in 1990 to 486 225 in 2021, while DALYs rose from 9.44 million to 11.81 million. The age-standardized mortality rate (ASMR) decreased from 8.8 to 5.6 per 100 000 [EAPC = −1.54], and the age-standardized DALY rate (ASDR) decreased from 228.5 to 134.7 per 100 000 [EAPC = −1.8]. In 2021, esophageal, stomach, and pancreatic cancers contributed the most deaths. The highest population-attributable fraction (PAF) for SADC deaths was observed in East Asia, High-Income North America, Central Europe, and Eastern Europe. China and Mongolia bear the greatest burden, with China having the highest number of deaths and DALYs globally, and Mongolia having the highest ASMR and ASDR, with increasing trends.

Conclusion:

Despite the downward trends in ASMR and ASDR, the global burden of SADC remains significant, with notable regional disparities. Effective anti-smoking policies and increased public awareness of smoking-related digestive cancer risks are essential to reduce smoking rates and mitigate the burden of SADC.

HIGHLIGHTS

• Global Burden: SADC deaths rose from 347 533 in 1990 to 486 225 in 2021, with DALYs increasing from 9.44 million to 11.81 million.

• Regional Disparities: Highest PAF in East Asia, North America, and Europe. China and Mongolia have the highest burden.

• Policy and Awareness: Effective anti-smoking policies and public awareness are crucial to reduce smoking rates and SADC burden.

Introduction

Digestive cancers rank among the most prevalent malignancies globally, accounting for 24% of all newly diagnosed cancer cases and 33.3% of cancer-related deaths worldwide[1]. The burden of these cancers is particularly pronounced in low-and middle-income countries (LMICs), where limited healthcare infrastructure and insufficient cancer prevention and screening programs contribute to delayed diagnoses and poorer outcomes[2]. Despite advancements in early detection and therapeutic interventions, the incidence and mortality rates of many digestive cancers remain high, driven by aging populations, shifts in lifestyle behaviors, and persistent disparities in access to healthcare services. As highlighted by the International Agency for Research on Cancer, smoking is a well-established cause of cancers affecting multiple organs, including the lungs, stomach, colorectum, liver, pancreas, and kidneys[3].

In 2019, an estimated 1.14 billion people worldwide smoked, consuming approximately 7.41 trillion cigarettes. That same year, tobacco use was responsible for 7.69 million deaths and the loss of over 200 million disability-adjusted life years (DALYs), representing the leading risk factor for men mortality and accounting for 20.2% of all men deaths globally[4]. Alarmingly, nearly 155 million individuals aged 15–24 used tobacco in 2019, with 82.6% initiating smoking between ages 14 and 25, underscoring the urgent need for targeted prevention efforts among adolescents and young adults[5].

However, substantial regional disparities in smoking trends persist across countries. Between 1990 and 2019, the most substantial increases in the number of smokers occurred in North Africa and the Middle East (104.1%) and sub-Saharan Africa (74.6%)[4]. Given these shifting smoking patterns and the growing global burden of digestive cancers, it is critical to quantify the impact of smoking on digestive cancer outcomes. This study aims to comprehensively estimate the burden of smoking-attributable digestive cancers (SADC) at the global, regional, and national levels. Our findings, including sex-specific analyses, can inform tobacco control policies and guide targeted cancer prevention strategies worldwide.

Materials and methods

Data source

This study is based on the Global Burden of Disease (GBD) 2021 study, which integrated a total of 100 893 data sources to estimate the burden of 371 diseases and injuries across 204 countries and 811 subnational regions. Compared with the GBD 2019 iteration, GBD 2021 incorporated 19 189 new data sources for disability-adjusted life years (DALYs), added 12 new causes, and introduced several important methodological updates[6]. Estimates of mortality and DALYs for smoking-attributable digestive cancers (SADC) – including esophageal, gastric, colorectal, liver, and pancreatic cancers – were extracted from the publicly accessible GBD Results Tool (https://vizhub.healthdata.org/gbd-results/). Data were obtained at the global, regional (21 GBD regions), and national levels, with corresponding 95% uncertainty intervals (UIs). Disease classifications adhered to the International Classification of Diseases, 10th Revision (ICD-10). The specific codes used for inclusion were as follows: esophageal cancer: C15.0–C15.9, D00.1, D13.0; gastric cancer: C16; colorectal cancer: C18–C21.9, D01.0–D01.3, D12–D12.9, D37.3–D37.5; liver cancer: C22–C22.4, C22.7–C22.8, Z85.05; pancreatic cancer: C25–C25.9. The work has been reported in line with the strengthening the reporting of cohort, cross-sectional, and case-control studies in surgery (STROCSS) criteria[7].

Data handling and modeling approach

Our study relies on data provided by GBD, which has already been processed using Bayesian models. The Bayesian approach combines observed data, prior probabilities, and the likelihood function to obtain the posterior probability[8]. For regions with sparse data, the model can utilize additional information from the prior distribution of similar regions to provide more reasonable estimates for sparse data, helping the model make informed inferences when data is limited. Bayesian models often handle cross-regional data variability through hierarchical random effects. Hierarchical models allow data from different regions to share certain parameters, while simultaneously allowing each region to retain its own specificity. This structure enables the model to capture commonalities across regions while reflecting the unique differences of each region. A key feature of Bayesian models is their ability to quantify uncertainty in parameter estimation. For sparse or highly variable data, Bayesian models provide a natural way to express uncertainty through the posterior distribution. In regions with insufficient or highly fluctuating data, the model’s uncertainty in the output will be higher.

Statistical analysis

To comprehensively assess the burden of SADC, we extracted age-standardized mortality rates (ASMRs) and age-standardized DALY rates (ASDRs) per 100 000 population from 1990 to 2021. Age standardization was performed using the GBD world standard population to allow for fair comparisons across regions and time. We also analyzed the age-standardized prevalence, incidence, and years lived with disability (YLDs) rates for both sexes across global, regional, and national levels. The estimated annual percentage change (EAPC) was calculated to evaluate temporal trends in ASRs over time. The EAPC was derived by fitting a linear regression model to the natural logarithm of the ASR values, following the formula: $ln\left(ASR\right)=\alpha +\beta \text{cdotx}+\epsilon$ (where $x$represents the calendar year, and $\beta$ is the slope of the regression line). The EAPC and its 95% confidence interval (CI) were then computed using the formula: $EAPC=100\times \left(exp\left(\beta \right)-1\right)$. A positive EAPC with a lower CI bound >0 indicates a statistically significant increasing trend, whereas a negative EAPC with an upper CI bound <0 suggests a significant decreasing trend[9]. Analyses were stratified by sex and location to explore disparities across subgroups. All statistical analyses and data visualizations were conducted using R software (version 4.3.3). A two-tailed P value <0.05 was considered statistically significant.

Results

Global burden of SADC

In 2021, an estimated 486 224.9 (95% uncertainty interval [UI]: 351 937.9–632 745.8) deaths from smoking-attributable digestive cancers (SADC) occurred worldwide, corresponding to a population-attributable fraction (PAF) of 13.8%. The global age-standardized mortality rate (ASMR) for SADC was 5.6 (95% UI: 4.1–7.3) per 100 000 population. The total number of disability-adjusted life years (DALYs) lost due to SADC was 11 811 096.1 (95% UI: 8 463 135.8–15 472 624), with an age-standardized DALY rate (ASDR) of 134.7 (95% UI: 96.5–176.4) per 100 000 population (Table 1). Among the five SADC subtypes, esophageal cancer had the highest PAF at 38.0% (95% UI: 30.8–44.6%), followed by stomach, pancreatic and liver cancers, all of which had PAFs exceeding 10% in 2021 (Fig. 1A). It also caused the most deaths (205 462.9; 95% UI: 156 544.4–257 209.4), exceeding those of stomach (107 925.6; 95% UI: 84 603.1–138 447.8), pancreatic (72 170; 95% UI: 62 852.9–82 937.5), liver (53 053.7; 95% UI: 18 267.6–88 111), and colorectal cancers (47 612.6; 95% UI: 29 669.9–66 040.2) (Fig. 2A-D). Smoking-attributable esophageal and stomach cancers remained the leading contributors to absolute DALYs throughout the study period (Fig. 2B). Pancreatic cancer surpassed liver and colorectal cancers in burden during 1990–2021, reflecting its growing proportion of SADC-related DALYs. Despite the rising number of deaths and DALYs associated with all SADC subtypes, the ASMR and ASDR for smoking-attributable esophageal and stomach cancers exhibited a marked decline over the past three decades (Fig. 2C, D; Supplementary Digital Content, Figure S1, available at: http://links.lww.com/JS9/F21).

Table 1.

Global burden of SADC by cancer category, 2021

		Deaths		DALYs
Cancer type	Sex	Smoking-attributable cases	Age-standardized rate (per 100 000 person-years)	Smoking-attributable cases	Age-standardized rate (per 100 000 person-years)
SADC	Both	486 224.9 (351 937.9 to 632 745.8)	5.6 (4.1 to 7.3)	11 811 096.1 (8 463 135.8 to 15 472 624)	134.7 (96.5 to 176.4)
	Female	47 230.4 (33 362.1 to 62 733.3)	1 (0.7 to 1.3)	1 038 188.5 (743 043.8 to 1 360 754.8)	22.5 (16.1 to 29.5)
	Male	438 994.5 (317 310.8 to 573 781)	11.1 (8 to 14.4)	10 772 907.6 (7 682 318.3 to 14 173 945.7)	259.2 (185.1 to 340.4)
Esophageal cancer	Both	205 462.9 (156 544.4 to 257 209.4)	2.4 (1.8 to 3)	4 765 032 (3 626 147.9 to 5 973 195.2)	54.3 (41.3 to 68)
	Female	12 181.9 (8538.9 to 16 514.6)	0.3 (0.2 to 0.4)	240 944.6 (169 832.7 to 318 850.7)	5.2 (3.7 to 6.9)
	Male	193 281.1 (147 756.8 to 241 824.8)	4.9 (3.7 to 6.1)	4 524 087.4 (3 441 582.8 to 5 667 126.1)	109.3 (83.2 to 136.8)
Stomach cancer	Both	107 925.6 (84 603.1 to 138 447.8)	1.3 (1 to 1.6)	2 537 997.7 (1 991 161.3 to 3 270 229.2)	29 (22.7 to 37.3)
	Female	7604.1 (6002.5 to 9441.1)	0.2 (0.1 to 0.2)	164 781.1 (132 120.5 to 203 386.1)	3.6 (2.9 to 4.4)
	Male	100 321.5 (78 097 to 129 829.8)	2.6 (2 to 3.3)	2 373 216.7 (1 848 951.9 to 3 077 861.1)	57.6 (44.8 to 74.6)
Pancreatic cancer	Both	72 170 (62 852.9 to 82 937.5)	0.8 (0.7 to 1)	1 789 502.8 (1 567 220.8 to 2 042 056.6)	20.4 (17.8 to 23.2)
	Female	14 909.3 (12 685.6 to 17 190.4)	0.3 (0.3 to 0.4)	334 618.6 (291 427.6 to 377 986.7)	7.2 (6.3 to 8.2)
	Male	57 260.7 (49 810.4 to 66 081.3)	1.4 (1.2 to 1.7)	1 454 884.3 (1 261 445.2 to 1 677 719.5)	34.7 (30.1 to 40)
Liver cancer	Both	53 053.7 (18 267.6 to 88 111)	0.6 (0.2 to 1)	1 482 896.3 (504 999.6 to 2 478 905.6)	16.9 (5.8 to 28.3)
	Female	4887.1 (1464.1 to 8775.5)	0.1 (0 to 0.2)	111 904.7 (33 712.7 to 200 080.7)	2.4 (0.7 to 4.3)
	Male	48 166.5 (16 663.7 to 80 568.2)	1.2 (0.4 to 2)	1 370 991.6 (467 666.5 to 2 286 527.4)	32.3 (11 to 54.1)
Colorectal cancer	Both	47 612.6 (29 669.9 to 66 040.2)	0.6 (0.3 to 0.8)	1 235 667.3 (773 606.3 to 1 708 237.4)	14.1 (8.8 to 19.5)
	Female	7648 (4670.9 to 10 811.7)	0.2 (0.1 to 0.2)	185 939.5 (115 950.4 to 260 450.7)	4.1 (2.5 to 5.7)
	Male	39 964.7 (24 983 to 55 476.9)	1 (0.6 to 1.4)	1 049 727.7 (662 671.9 to 1 464 711.6)	25.2 (15.9 to 35)

DALYs: disability-adjusted life years; SADC: smoking-attributable digestive cancers.

Figure 1.

Population attributable fractions, by cancer category and region: (A) both sexes combined; (B) men; (C) women; SADC: smoking-attributable digestive cancers.

Figure 2.

Global burden trend of SADC stratified by cancer category: (A) all age counts; (B) age-standardized rates. (C)ASR: age-standardized rates; (D)SADC: smoking-attributable digestive cancers. It should be noted that Figure 2 presents the overall burden of digestive cancers, whereas the numerical values reported in the text specifically refer to smoking-attributable digestive cancers (SADC). Therefore, the ranking of cancer types by death counts and DALYs differs between the two, although the overall trends remain consistent.

Region burden of SADC

Figure 3 and Supplementary Digital Content, Figure S2 (available at: http://links.lww.com/JS9/F21) demonstrate significant regional variations in SADC burden. Despite declining ASMR and ASDR from 1990–2021, East Asia maintained the highest SADC burden in 2021, with peak mortality (274 548.8 deaths; 95% UI: 191 064–376 850), DALYs (11 811 096.1; 95% UI: 8 463 135.8–15 472 624), ASMR (12.5 per 100 000; 95% UI: 8.7–17.1), and ASDR (295.7 per 100 000; 95% UI: 202.3–410.9). Additionally, Western Europe, High-Income North America, and High-Income Asia Pacific also experienced a substantial burden, with more than 20 000 deaths and over 500 000 DALYs due to SADC in 2021. In contrast, Oceania and Central Sub-Saharan Africa reported the lowest absolute numbers of SADC-related deaths and DALYs among the 21 GBD regions, while Western Sub-Saharan Africa recorded the lowest ASMR and ASDR (Fig. 3; Supplementary Digital Content, Figure S2, available at: http://links.lww.com/JS9/F21; Supplementary Digital Content, Table S1, available at: http://links.lww.com/JS9/F21). Overall, while the absolute number of deaths and DALYs increased across nearly all regions over the past 32 years, ASDR declined to varying extents (Supplementary Digital Content, Table S1). In each region, esophageal cancer consistently exhibited the highest PAF among the five SADC subtypes, with pancreatic cancer ranking second (Fig. 1A). In 2021, the PAF for smoking-attributable esophageal cancer exceeded 40% in East Asia (46.7%), High-Income North America (46.0%), High-Income Asia Pacific (41.4%), and Eastern Europe (41.1%). The patterns of mortality among SADC subtypes slightly differed from those of the PAFs; smoking-attributable esophageal cancer caused the most deaths, followed by stomach cancer. In East Asia, High-Income Asia Pacific, Western Europe, and Eastern Europe, ASMR and ASDR for each SADC type remained relatively high (Supplementary Digital Content, Figure S2, available at: http://links.lww.com/JS9/F21), though the magnitude of these metrics varied by cancer type. Detailed region-specific estimates stratified by cancer type are presented in Supplementary Digital Content, Tables S1–S6 (available at: http://links.lww.com/JS9/F21).

Figure 3.

Region burden trends of SADC from 1990 to 2021: (A) all age counts (deaths cases); (B) all age counts (DALYs). SADC: smoking-attributable digestive cancers; DALY: disability-adjusted life years.

Country burden of SADC

From 1990 to 2021, the number of deaths from SADC doubled or more in 52 out of 204 countries and territories, with 168 countries experiencing an overall increase. Similarly, DALYs doubled or more in 51 countries, and increased in 162. Among them, China, the United States, and Japan reported the highest total number of all-age SADC deaths in 2021 (Supplementary Digital Content, Table S7). At the national level, Mongolia exhibited the highest ASMR (16.6 per 100 000; 95% UI: 8.6–25.3), followed by China (12.7 per 100 000; 95% UI: 8.8–17.4) and Denmark (Greenland) (11.2 per 100 000; 95% UI: 7.5–15.9). Corresponding ASDR values were 444.7 (95% UI: 226.8–684.8), 299.3 (95% UI: 203.3–417.3), and 285.1 (95% UI: 194.3–397.4) per 100 000, respectively (Fig. 4A, B).

Figure 4A.

Country trends of SADC from 1990 to 2021: (A) age-standardized mortality rate in 2021; (B) age-standardized DALYs rate in 2021; (C) EAPC of ASMR from 1990 to 2021; (D) EAPC of ASDR from 1990 to 2021, SADC: smoking-attributable digestive cancers; DALYs: disability-adjusted life years; EAPC: estimated annual percentage change.

FIGURE 4B.

For ASMR, 186 countries showed a decline, with Turkmenistan exhibiting the most substantial reduction – from 13.9 (95% UI: 11.1–17.0) to 3.9 (95% UI: 2.5–5.5) per 100 000 – and an EAPC of −4.07 (95% CI: −4.44 to −3.70). In contrast, Lesotho had the greatest increase, with ASMR rising from 4.9 (95% UI: 2.8–7.6) to 8.8 (95% UI: 5.1–15.5) per 100 000, and an EAPC of 2.50 (95% CI: 2.13–2.86) (Supplementary Digital Content, Table S7, available at: http://links.lww.com/JS9/F21; Fig. 4A, C). A similar pattern was observed for ASDR, which declined in 185 countries. The most marked reduction occurred in the Republic of Korea, with ASDR decreasing from 543.3 (95% UI: 331.4–762.0) to 140.8 (95% UI: 84.2–203.9) per 100 000, and an EAPC of −4.71 (95% CI: −4.83 to −4.60). Notable decreases were also observed in Turkmenistan (EAPC = −4.19; 95% CI: −4.53 to −3.84) and Singapore (EAPC = −3.67; 95% CI: −3.77 to–3.57). Lesotho again exhibited the greatest increase in ASDR, rising from 116.0 (95% UI: 67.2–181.0) to 235.9 (95% UI: 130.8–428.5) per 100 000, with an EAPC of 2.88 (95% CI: 2.46–3.29) (Supplementary Digital Content, Table S7, available at: http://links.lww.com/JS9/F21; Fig. 4B, D). Among the five SADC subtypes, esophageal and gastric cancers consistently accounted for the highest mortality and DALY burdens at the national level. However, the burden patterns of the five SADC subtypes vary across countries. Country–specific estimates by cancer type are available in Supplementary Digital Content, Tables S7–S12 (available at: http://links.lww.com/JS9/F21).

Burden of SADC by sex and age group

In 2021, the PAF of SADC–related deaths were markedly higher in men than in women, a disparity consistently observed across all subtypes of SADC (Fig. 1B, C). That year, the number of deaths attributable to SADC was 438 994.5 (95% UI: 317 310.8 573 781) in men, compared to 47 230.4 (95% UI: 33 362.1–62 733.3) in women. The corresponding PAFs for men and women was 19.9% and 3.3%, respectively. Similarly, the DALYs were substantially higher in men – 10 772 907.6 (95% UI: 7 682 318.3–14 173 945.7) – versus 1 038 188.5 (95% UI: 743 043.8–1 360 754.8) in women. This sex-based disparity in disease burden persisted across all five SADC subtypes (Table 1). The ASMR for SADC in 2021 was 11.1 (95% UI: 8.0–14.4) for men, compared to 1 (95% UI: 0.7–1.3) per 100 000 for women. Likewise, the ASDR for SADC was 259.2 (95% UI: 185.1–340.4) for men, while it was only 22.5 (95% UI: 16.1–29.5) per 100 000 for women. This pattern of significantly higher rates in men was consistent across all five cancer subtypes (Table 1).

Across all age groups, men exhibited higher mortality and DALY rates of SADC than women, following a trend of increasing burden with age until late adulthood, followed by a subsequent decline. Higher mortality rates were predominantly observed in the 70 years and older age group, while higher DALY rates were concentrated among 60–79-year-olds. Specifically, the highest mortality rate occurred in the 85–89 age group, reaching 66.1 (95% UI: 47.1–87.0), while the peak DALY rate was observed in the 70–74 age group at 888.9 (95% UI: 643.7–1169.5) (Fig. 5A, B). In terms of absolute numbers, deaths and DALYs were consistently higher in men than in women across all age groups, with the greatest burden observed among individuals aged 55–79. The highest number of deaths and DALYs occurred in the 65–69 age group, with 90,061.1 (95% UI: 64 579.2–117 186.5) deaths and 2 225 782.0 (95% UI: 1 594 678.5–2 895 810.6) DALYs, respectively (Fig. 5A, B). Between 1990 and 2021, most SADC-related deaths occurred in the 55–79 age group. Globally, there has been a notable epidemiological transition, with the burden of SADC gradually shifting from younger age groups (<55 years) to older populations (>75 years) over the past three decades (Fig. 5C, D).

Figure 5.

The global trend of SADC deaths and DALYs based on gender and age groups: (A) death cases and mortality rate of SADC per 100 000 person-years by age and sex, 2021; (B) DALYs and DALYs rate of SADC per 100 000 person-years by age and sex, 2021; (C) death cases of SADC by age from 1990 to 2021; (D) death percentage of SADC by age from1990 to 2021, SADC: smoking-attributable digestive cancers; DALYs: disability-adjusted life years.

Discussion

In 2021, an estimated 486 000 deaths from digestive cancers – accounting for 13.8% of all such deaths – were attributable to smoking, with approximately 90% of these deaths occurring in men. The highest PAFs were observed in East Asia, High-Income North America, and Central Europe, whereas the lowest were recorded in Western Sub-Saharan Africa and Andean Latin America. Among the 204 countries and territories assessed, China reported the highest absolute number of deaths and DALYs due to SADC, while Mongolia exhibited the highest ASMR and ASDR[6]. These findings highlight the urgent need for regionally tailored public health interventions, particularly in high-burden areas such as East Asia and Central Europe. Effective strategies may include stricter tobacco taxation, large-scale health education campaigns, and improved enforcement of smoke-free legislation^[10–12]. Tailoring interventions to the specific cultural, economic, and healthcare infrastructure of each region is essential to achieving sustained reductions in SADC burden[13].

To better understand the context of this burden, it is essential to examine historical trends in smoking prevalence, which directly influence disease patterns over time. According to the World Health Organization’s Global Report on Trends in Prevalence of Tobacco Smoking 2000–2025, the global smoking rate has declined steadily, decreasing from 27% in 2000 to 20% in 2015, and is projected to further fall to 17.2% by 2025. Despite this overall downward trajectory, substantial regional and national variations in smoking prevalence persist[14]. The 2023 WHO Report on the Global Tobacco Epidemic confirms a consistent reduction in smoking rates, particularly among adults aged 15 years and older. Nonetheless, over one billion individuals worldwide continue to smoke, representing a major ongoing threat to global public health[15]. Moreover, the global decline in smoking rates can largely be attributed to the continued implementation of comprehensive tobacco control policies across countries, which has contributed to the observed reductions in SADC–related mortality and DALY rates. The MPOWER strategy, launched by the World Health Organization in 2008, is widely recognized as a key policy framework driving this downward trend. MPOWER comprises six core components, including tobacco use monitoring, protection from tobacco smoke exposure, cessation support, health warnings, bans on tobacco advertising, and increased tobacco taxation[10]. Several countries, such as the United States and Australia, adopted multiple components of the strategy early on, including the implementation of graphic health warnings, comprehensive public smoking bans, and systematic cessation intervention services, all of which significantly accelerated the reduction in smoking–related cancer burden^[16,17]. Studies have shown that following the implementation of robust tobacco control policies, the decline in ASMR in the United States significantly exceeded the global average, with Australia achieving similar progress. These findings suggest that comprehensive and systematic tobacco control policies play a critical role in reducing the burden of SADC, particularly in countries and regions with high policy coverage and strong enforcement. While countries like the United States and Australia have achieved significant progress through comprehensive adoption of MPOWER, other regions continue to face substantial challenges. These discrepancies in implementation and enforcement contribute to the persistent geographical disparities in SADC burden observed worldwide. Nevertheless, pronounced geographical disparities in SADC burden remain. In 2021, East Asia bore the greatest burden of SADC, yet it also experienced the most significant reductions in both ASMR and ASDR between 1990 and 2021. In contrast, Western Sub-Saharan Africa had the lowest SADC burden, with the lowest PAF, ASMR, and ASDR. These differences are likely influenced by varying levels of implementation and enforcement of tobacco control policies, as well as persistently high smoking prevalence in certain regions, particularly in parts of Asia[18].

Among high-burden countries, Mongolia presents a particularly complex scenario. Despite implementing various tobacco control measures, including increased taxation and public smoking bans, the country continues to report rising ASMR and ASDR for SADC. These trends suggest that partial implementation and limited enforcement may hinder the effectiveness of current policies. Although Mongolia has implemented several tobacco control measures – such as increased tobacco taxation, smoke-free legislation, and public health education campaigns – the effectiveness of these policies has been significantly constrained by limited enforcement and poor public compliance. Moreover, the country lacks a structured smoking cessation support system, which further compromises the impact of its tobacco control efforts. Mongolia also remains a high-burden area for chronic hepatitis B and C infections, and the synergistic effects between viral hepatitis and smoking may have exacerbated the incidence and burden of liver cancer and other digestive system malignancies. Compared to countries that have fully adopted the WHO MPOWER strategy – such as the United States and Australia – Mongolia’s partial implementation may help explain the continued rise in age-standardized mortality rate (ASMR) and disability-adjusted life years (ASDR) related to SADC observed in 2021^[10,19,20]. These findings suggest that Mongolia faces critical challenges in improving the enforcement of tobacco control legislation, developing a comprehensive cessation infrastructure, and integrating strategies that simultaneously address comorbid conditions, such as chronic viral hepatitis.

In contrast to Mongolia’s relatively small population but high per-capita burden, China bears the largest absolute burden of SADC deaths and DALYs globally. Although its age-standardized rates are lower than Mongolia’s, China’s vast smoking population – coupled with cultural norms and economic ties to tobacco production – continues to sustain a significant public health challenge. China accounted for the highest number of SADC deaths and DALYs, exceeding those of the United States – the second-ranked country – by more than tenfold. Although China’s ASMR and ASDR were second only to Mongolia, its overall burden remains substantial. Despite recent declines in smoking rates, China retains the largest smoking population worldwide, driven largely by its massive population base. While multiple tobacco control policies have been introduced, their impact has been constrained by entrenched cultural norms and economic dependencies related to the tobacco industry[15]. These findings emphasize the persistent and significant residual burden of SADC mortality and DALYs, underscoring the urgent need for intensified, context-specific tobacco control efforts.

To complement the population-level analysis, it is also important to consider the biological mechanisms through which smoking drives digestive cancers. A clearer understanding of these pathways can further justify targeted intervention, particularly for cancers like esophageal carcinoma that exhibit the highest smoking-attributable fraction. Smoking contributes to the development of digestive cancers through multiple biological mechanisms. These include the induction of chronic inflammation, the formation of carcinogenic compounds such as nitrosamines, and the promotion of DNA damage and mutations^[20–23]. Moreover, smoking can undermine immune surveillance and disrupt normal cellular processes[24], thereby heightening the risk of malignant transformation in digestive tissues. Our findings indicate that among the five subtypes of SADC, esophageal cancer accounted for the highest number of deaths and DALYs, as well as the highest PAF. This further reinforces that smoking is the most significant modifiable risk factor for esophageal cancer. Individuals with the ALDH2 functional variant risk genotype have an inherently higher baseline risk of developing esophageal cancer, a risk that is further exacerbated by lifestyle factors like smoking and alcohol consumption[25]. It is estimated that 28% to 45% of the East Asian population carries the ALDH2 risk genotype[26]. Consequently, in our study, a proportion of esophageal cancer cases linked to smoking among East Asians may be due to the heightened risk associated with this genetic variant. Studies have shown that smoking doubles the risk of Barrett’s esophagus patients progressing to cancer or high-grade dysplasia compared to those who have never smoked[27]. The highest PAF of smoking-attributable esophageal cancer is not only linked to the crucial role smoking plays in the onset of the disease but also to its role in promoting the progression of esophageal cancer. Nicotine, a key component of tobacco smoke, has been shown to strongly upregulate the RNA-binding protein ILF2 through the activation of the JAK2/STAT3 signaling pathway. ILF2 binds to the THO complex protein THOC4, facilitating the selective nuclear export of mRNAs encoding pluripotency transcription factors SOX2, NANOG, and SALL4. This process increases the nuclear export and stability of these stemness factor mRNAs, thereby enhancing the cancer stem cell properties and chemoresistance of esophageal cancer cells[28]. Experimental studies indicate that nicotine downregulates the deubiquitinating enzyme OTUD3. Normally, OTUD3 stabilizes the RNA-binding protein ZFP36, which promotes the rapid degradation of VEGF-C mRNA. Nicotine-induced downregulation of both OTUD3 and ZFP36 results in elevated VEGF-C expression, thereby facilitating lymphatic metastasis[29]. We believe that the PAF for esophageal cancer mortality is the highest among the five SADC subtypes, not only because smoking first contacts and damages the esophageal mucosa, but also due to its critical role in the progression of esophageal cancer.

Across the global, regional, and national levels, men bear a heavier SADC burden compared to women. This observed phenomenon is not only related to physiological differences between genders but is primarily attributed to the significantly higher number of men smokers compared to women smokers. According to the latest report, in 2021, the global smoking rate was 29% for men and only 5% for women. Over the past 15 years, smoking rates have decreased for both genders worldwide, with a more pronounced decline among women[15]. In 2021, the global death rate of SADC generally increased with age, except for a slight decline observed in the 95+ age group compared to the 90–94 age group. The high mortality rate of SADC in elderly populations may be attributed to the cumulative effects of long-term smoking, which lead to the accumulation of genetic mutations and a decline in immune system function. From 1990 to 2021, the age distribution of SADC mortality shifted from the under-54 age group to those over 75. This trend indicates that, over time, changes in social, environmental, and demographic factors have profoundly influenced the composition of the global SADC burden. The rise in average life expectancy and the increasing proportion of the elderly population are key factors driving the growing burden of age-related SADC.

Despite a certain degree of public cognizance concerning the correlation between smoking and the risk of cancer, a considerable population continues to uphold their smoking habits on a global scale, particularly among men. The World Tobacco Control Report (2021) asserts that elevating tobacco taxes can markedly diminish tobacco consumption, especially in low- and middle-income nations[30]. A growing array of countries and regions are instituting expansive smoke-free areas, encompassing public spaces, workplaces, and dining establishments. The Global Tobacco Surveillance Report (2023) reveals that such thorough smoke-free policies have significantly reduced exposure to secondhand smoke and have inspired a greater number of smokers to cease their habit[15]. Nevertheless, global tobacco control continues to face challenges. In low- and middle-income countries, the enforcement and scope of tobacco control laws remain inadequate. In certain regions and nations, smoking behaviors are often influenced by cultural and social norms, necessitating prolonged education and advocacy to shift these entrenched perceptions^[31,32]. By gaining insights into smoking behaviors and effective intervention measures within different countries and cultural contexts, conducting large-scale health education and awareness campaigns in developing nations, and implementing comprehensive smoking bans, we can significantly reduce the burden of SADC.

The COVID-19 pandemic likely disrupted ongoing tobacco control efforts across many countries. When evaluating the disease burden of SADC from 1990 to 2021, it is reasonable to consider the impact of the COVID–19 pandemic on mortality. Although this study systematically evaluated the burden of smoking–attributable digestive cancers (SADC) at the global, regional, and national levels from 1990 to 2021, it did not account for the potential impact of the COVID–19 pandemic. Between 2020 and 2022, public health resources in many countries were extensively redirected toward pandemic response efforts, leading to the suspension or deprioritization of tobacco control programs[33]. Disruptions in smoking cessation clinics, helpline services, and the enforcement of tobacco-related regulations may have indirectly influenced smoking behaviors and intensity[34]. While current datasets have not yet fully captured the pandemic’s effect on smoking prevalence and its long-term implications for SADC, existing evidence suggests that this policy gap could have considerable delayed consequences on global SADC mortality and disability-adjusted life years (DALYs) in the coming decades[35]. Future research should prioritize longitudinal evaluations of how interruptions in tobacco control interventions during the COVID–19 era may affect the burden of SADC over time. The COVID–19 pandemic may have also impacted individuals’ overall smoking consumption and altered smokers’ habits. Despite disruptions in data collection from 2020 to 2022 due to the pandemic in many countries, studies from China, Italy, the United States, and other nations have found a weakening in smoking cessation activity. Greater efforts are needed to enhance motivation to quit^[36–38]. However, concerning mortality and DALYs of SADC, changes in individual smoking patterns during the COVID–19 pandemic have not yet become apparent, but they may be reflected in the coming decades.

A key limitation of this study lies in the potential confounding effects of co-existing risk factors, particularly in the estimation of liver cancer attributable to smoking. This issue is especially relevant in countries with a high prevalence of viral hepatitis, such as Mongolia, where approximately 44% and 46% of liver cancer cases are attributable to hepatitis B and C virus infections, respectively^[39,40]. Smoking may act synergistically with HBV/HCV infections to promote hepatocarcinogenesis through mechanisms such as chronic inflammation, DNA damage, and immune suppression. In such high-burden settings, the combined effects of smoking and viral infections may lead to over- or underestimation of the smoking-attributable burden. Moreover, the current GBD estimation methods have limited capacity to adjust for these comorbid factors, highlighting the need for future studies to incorporate multiple risk factor modeling to improve the accuracy and policy relevance of attribution estimates.

Conclusions

In summary, despite various measures implemented by authorities to reduce the number of smokers, the deaths and DALYs of SADC have shown a significant increase over the course of the past 32 years. Esophageal cancer holds the highest PAF of SADC, accounting for approximately half of the SADC’s mortality and DALYs. The Eastern Asian countries of China and Mongolia bear a substantial burden of SADC. Strict adherence to the WHO Framework Convention on Tobacco Control, coupled with region-specific interventions targeting to the distinct smoking patterns, can effectively prevent a significant proportion of digestive cancers.

Ethical approval

Not applicable.

Consent

Not applicable.

Sources of funding

This study was financially supported by National Natural Science Foundation of China (No. 81972269).

Author contributions

M.Z., Z.X., B.L., and W.Y.: methodology, formal analysis, writing – original draft; D.W.: design improvement, administrative and material support, supervision.

Conflicts of interest disclosure

The authors declare no conflicts of interest.

Research registration unique identifying number (UIN)

None.

Guarantor

Daorong Wang.

Provenance and peer review

Not commissioned, externally peer-reviewed.

Data availability statement

The datasets used during the current study are available from the corresponding author upon reasonable request.