Burden, trends and predictions of esophageal cancer in the top five GDP countries: a systematic analysis based on the Global Burden of Disease 2023 study

Abstract

Background

Esophageal cancer remains a major global health burden, but comparative analyses across leading economies are limited. This study aimed to comprehensively evaluate and compare the esophageal cancer burden from 1990 to 2023 and project trends to 2050 in the top five gross domestic product (GDP) countries-China, Germany, India, Japan, and the United States-using data from the Global Burden of Disease (GBD) 2023 study.

Methods

We extracted age-, sex-, and risk factor-specific data on esophageal cancer incidence, mortality, and disability-adjusted life years (DALYs) for the five countries from the GBD 2023 database. Age-standardized rates (ASRs) were calculated, and temporal trends were assessed using Joinpoint regression. Age-period-cohort (APC) modeling was employed to disentangle the independent effects of age, period, and cohort. Decomposition analysis quantified the contributions of population aging, population growth, and epidemiological changes. Bayesian age-period-cohort (BAPC) models were applied to project disease burden through 2050.

Results

China had the highest absolute burden but achieved the steepest declines. India’s age-standardized incidence rates (ASIRs) more than doubled. Japan, Germany, and the USA maintained lower, stable rates. APC analysis showed an “inverted-U” age peak in China versus older peaks elsewhere. Decomposition identified epidemiological changes as the main alleviating factor in China, whereas aging and unfavorable trends drove India’s rise. Smoking dominated in four countries; India was co-driven by chewing tobacco and low vegetable intake. BAPC projections foresee continued declines in China, Germany, and Japan, a surge in India, and a rebound in the USA.

Conclusions

The esophageal cancer burden exhibits marked heterogeneity across the top five GDP countries, driven by distinct demographic transitions, risk factor profiles, and public health responses. These findings emphasize the urgent need for tailored, context-specific cancer control strategies, including strengthened tobacco and alcohol regulation, dietary improvements, and expanded screening in high-risk populations.

Highlight box.

Key findings

• Using Global Burden of Disease (GBD) 2023 data, China had the highest esophageal cancer burden but steepest declines; India’s incidence rate doubled. Other three countries remained stable. Smoking was the dominant risk factor in four nations; India was jointly driven by chewing tobacco and low vegetable intake. Decomposition showed favorable epidemiological changes in China, but unfavorable in India. Bayesian age-period-cohort (BAPC) model projects continued declines in China, Germany, and Japan, a surge in India, and a rebound in the USA.

What is known and what is new?

• Esophageal cancer is a major global malignancy with geographic and economic disparities; tobacco, alcohol, and dietary factors are key risk factors; high-income countries generally have lower rates.

• This is the first comparative, multi-method analysis of the five largest gross domestic product (GDP) economies using GBD 2023 data, integrating Joinpoint analysis, age-period-cohort (APC) model, decomposition analysis, and BAPC model to quantify age, period, cohort effects and project 2050 trends. It uncovers distinct transition patterns and risk-factors profiles across these countries.

What is the implication, and what should change now?

• Tailored, context-specific strategies are urgently needed. All countries must prepare for aging-driven increases in absolute case numbers by enhancing healthcare system capacity and workforce training.

Introduction

Esophageal cancer remains one of the most common malignant tumors of the digestive system worldwide. According to the latest GLOBOCAN 2022 data, esophageal cancer accounts for 2.6% of all new cancer cases and 4.6% of all cancer-related deaths, ranking as the eleventh most frequently diagnosed cancer and the seventh leading cause of cancer mortality globally (1). Early symptoms are often nonspecific and insidious, frequently resulting in delayed medical consultation. By the time overt manifestations such as progressive dysphagia, retrosternal pain, or unintentional weight loss appear, the disease is typically already at a locally advanced or metastatic stage. This late diagnosis severely constrains therapeutic options and contributes to persistently poor survival outcomes (2,3). Diagnostic delay thus represents a critical factor contributing to the high case-fatality ratio observed in esophageal cancer (4). Furthermore, the multimodal treatment regimens required, including surgery, radiotherapy, chemotherapy, and immunotherapy, are resource-intensive, imposing a significant financial burden on both patients and healthcare systems worldwide (5,6).

Geographic and economic disparities in esophageal cancer burden are pronounced. High-incidence areas stretch from eastern Asia to eastern and southern Africa, whereas Western countries have witnessed a rising incidence of adenocarcinoma over recent decades (7,8). Although previous studies have described global or regional epidemiological patterns, few have systematically compared the disease burden across the world’s largest economies using a unified data source (9). Such comparisons are essential because countries at different stages of economic development often face distinct mixtures of risk factors, screening capacities, and treatment accessibility.

Selecting representative countries for in-depth comparison helps reveal the mechanisms driving epidemiological shifts in esophageal cancer. Based on the World Bank’s World Development Indicators (https://databank.worldbank.org/source/world-development-indicators), the United States, China, Germany, Japan and India ranked as the top five economies by nominal gross domestic product (GDP) in 2024 and account for more than half of the global economy. Although these countries have significant economic influence, there are marked differences among them in terms of population structure, major risk factors (such as smoking, alcohol consumption, and dietary habits), and the capacity of their healthcare systems to provide screening and treatment. Consequently, the patterns of esophageal cancer incidence and mortality are influenced by these factors (10-12). Comparing these five nations within the Global Burden of Disease (GBD) framework allows us to disentangle the roles of population aging, risk factor exposure, and health system responses. Moreover, as China and India continue to undergo rapid socioeconomic transitions, lessons from Japan, the United States, and Germany may inform future prevention and early detection strategies (13,14).

In this study, we extracted data from the GBD 2023 study to conduct a comparative analysis of esophageal cancer burden across the top five GDP countries from 1990 to 2023. We examined age-standardized incidence, mortality, and disability-adjusted life year (DALY) rates by sex and age group. Joinpoint regression was applied to identify temporal turning points, and decomposition analysis quantified the contributions of population aging, growth, and epidemiological change, with projections extended to 2050 via Bayesian age-period-cohort (BAPC) modeling. By highlighting cross-national similarities and differences, our findings aim to provide evidence-based recommendations for optimizing esophageal cancer control strategies, with particular emphasis on resource allocation and targeted screening in high-burden settings. We present this article in accordance with the GATHER reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-2274/rc).

Methods

Data sources

The data for this study were derived from the GBD 2023 study (https://vizhub.healthdata.org/gbd-results), a comprehensive and standardized assessment of health loss attributed to 375 diseases and injuries and 88 risk factors across 204 countries and territories from 1990 to 2023 (15). We extracted esophageal cancer incidence, mortality, and DALYs for the top five economies by gross domestic product—the United States, China, Japan, Germany, and India-using the GBD Results Tool (http://ghdx.healthdata.org/gbd-results-tool), with query parameters covering the period 1990–2023 and all available age groups for both sexes. Age-standardized rates (ASRs) were applied to account for differences in population age structures across datasets, thereby enhancing comparability (16). The GBD 2023 framework integrates diverse primary data sources, including vital registration systems, population-based cancer registries, household surveys, and published epidemiological literature, subject to rigorous quality control protocols (17). To ensure comparability across countries and over time, all rates were age-standardized to the GBD world population standard. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Because this study relied exclusively on de-identified aggregate data from publicly available repositories, institutional review board approval was not required. All statistical analyses were conducted using R software (version 4.2.2), and statistical significance was defined as P<0.05.

Joinpoint regression analysis

Joinpoint regression was employed to delineate long-term trends in esophageal cancer burden across the five countries from 1990 to 2023. Analyses were conducted using the Joinpoint Regression Program (version 5.0.2, National Cancer Institute, USA), which fits piecewise log-linear models to identify points where trends change significantly (18,19). The analysis allowed for a maximum of five turning points. Grid search combined with the Monte Carlo permutation test was used to identify the optimal number and location of turning points, with a significance level set at α=0.05. For each segment, the annual percentage change (APC) was estimated, and the average annual percentage change (AAPC) was derived as a weighted average of segment-specific APCs over the entire period. A trend was deemed statistically significant when the 95 % confidence interval (CI) for APC or AAPC excluded zero. All procedures adhered to standard protocols from the National Cancer Institute’s Surveillance Research Program (https://surveillance.cancer.gov/help/joinpoint).

Age-period-cohort (APC) analysis

To disentangle the independent contributions of age, time period, and birth cohort to esophageal cancer burden, we fitted an APC model using the method proposed by Holford to provide a clearer understanding of the disease trends across age groups, time periods, and birth cohorts (20,21). Age effects (α) capture variations in disease risk associated with biological aging and cumulative exposure over the life course. Period (β) effects reflect temporal influences-such as shifts in diagnostic practices, treatment advances, or screening policies-that affect all age groups simultaneously. Cohort effects (γ) represent differences in risk attributable to early-life exposures and lifestyle patterns specific to each generation. The APC model is mathematically expressed as follows:

$$Y_{ijk}=\mu +{\alpha }_i+{\beta }_j+{\gamma }_k+{ϵ}_{ijk}$$ [1]

In this model, Y_ijk denotes the observed burden of disease for age group i, period j, and cohort k. The parameter μ serves as the global intercept, and ϵ_ijk captures the random error component.

Net drift was defined as the overall annual percent change in log-transformed disease rates after adjusting for nonlinear period and cohort effects, characterizing the overall long-term secular trend of the disease burden. Local drift, conversely, was defined as the annual percent change in log-transformed rates within each individual age group, allowing precise identification of temporal heterogeneity in disease risk changes across age strata.

Decomposition analysis

We applied decomposition analysis to quantify the drivers underlying changes in esophageal cancer burden from 1990 to 2023 across the five countries. Following the Das Gupta framework, we partitioned the total change in incidence, deaths and DALYs into three components: population aging, population growth, and epidemiological changes (e.g., changes in age-specific rates). The equation involves:

$$\mathrm{\Delta }Y=\mathrm{\Delta }P+\mathrm{\Delta }A+\mathrm{\Delta }E$$ [2]

ΔY stands for the aggregate burden change. Within this decomposition, ΔP is the component driven by an expanding total population, ΔA is the portion explained by demographic aging, and ΔE is the residual part linked to changes in age-specific rates. This approach isolates the net contribution of each factor while holding the other two constant, with the sum of component-specific effects equaling the total observed change. Results are presented as absolute contributions and relative proportions, where positive values indicate a factor contributing in the same direction as the overall trend.

Risk factor analysis

To quantify the contribution of modifiable risk factors to esophageal cancer burden, we used the comparative risk assessment framework of GBD 2023. Population attributable fractions (PAFs) were extracted for smoking, high alcohol use, low vegetable intake, and chewing tobacco. The PAF represents the proportional reduction in disease burden that would occur if current exposure were reduced to the theoretical minimum risk exposure level (TMREL). PAFs were estimated using five inputs: the population exposure distribution and its standard deviation, relative risks for each risk-outcome pair, the TMREL, and epidemiological information on the non-independence of causal pathways among risks. This framework accounts for overlap and mediation among risk factors, avoiding double-counting when attributable burdens are aggregated. We present attributable deaths and DALYs for each of the five countries, with 95% uncertainty intervals (UIs). Detailed methodological descriptions have been published previously (22-24).

BAPC forecasting model

To forecast the future trajectory of esophageal cancer burden across the five countries, we employed a BAPC model (25). This approach extends the conventional APC framework by incorporating prior distributions for age, period, and cohort effects, which are updated with observed data to generate posterior predictive distributions. Independent second-order random walk (RW2) priors were assigned to the age, period, and cohort effects to reflect the expected smoothness of temporal trends. Parameter estimation was performed using integrated nested Laplace approximations (INLA), which provide efficient and accurate Bayesian inference for such hierarchical structures. Based on incidence and mortality data from the GBD database for 1990–2023, the projections to 2050 assumed that past age-specific trends in incidence, death, and DALYs would continue in the absence of major structural changes. UIs were derived from the posterior distributions to reflect the variability inherent in the forecasts. In addition, this study employed a rigorous out-of-sample backtesting design to validate forecast accuracy: the model was trained using only data from 1990–2013 and then used to perform genuine out-of-sample extrapolation for 2014–2023. Forecast uncertainty was quantified through posterior sampling and RW2 conditional extrapolation, and forecast accuracy was expressed as the mean absolute percentage error (MAPE).

Statistical analysis

All analyses were conducted using the R software (version 4.2.2), with the BAPC, INLA, segmented, forecast, ggplot2 R packages. Estimates of incidence, mortality, DALYs, and corresponding rates were reported with 95% uncertainty intervals (95% UIs) to reflect estimation uncertainty. A two-sided P value <0.05 was considered statistically significant.

Results

Change in esophageal cancer disease burden

Table 1 shows significant differences in the burden of esophageal cancer in China, Germany, India, Japan, and the United States between 1990 and 2023. Among the five countries, China consistently recorded the highest absolute numbers and ASRs across all measures in both years, though its rates declined substantially over the study period. The age-standardized incidence rate (ASIR) in China dropped from 28.23 per 100,000 in 1990 to 11.45 per 100,000 in 2023, with a corresponding decrease in age-standardized mortality rate (ASMR) from 28.46 to 10.86 per 100,000. In stark contrast, India experienced the most pronounced upward trajectory, with its ASIR more than doubling from 3.71 to 7.77 per 100,000 and its EAPC reaching 2.21% (95% CI: 1.64, 2.78) for incidence and 2.25% (95% CI: 1.68, 2.83) for mortality. The remaining three countries-Japan, Germany, and the United States-maintained relatively stable and lower rates, with age-standardized incidence in 2023 ranging from 4.15 to 6.21 per 100,000. Japan recorded the highest rate among them (6.21 per 100,000), followed by Germany (4.99 per 100,000) and the United States (4.15 per 100,000). Across all countries, the age-standardized DALYs rates (ASDR) followed similar patterns to incidence and mortality, with China showing the steepest decline and India the sharpest increase over the 33-year period.

Table 1. Changes in the disease burden of esophageal cancer in five countries: all-ages cases, age-standardized rates, and trends from 1990 to 2023.

Location	Measure	1990		2023		EAPC, 1990–2023 (95% CI)
		All-age cases (95% UI)	Age-standardized rate (95% UI)	All-age cases (95% UI)	Age-standardized rate (95% UI)
China	Incidence	243,228.75 (171,931.12, 285,757.87)	28.23 (20.29, 33.13)	267,431.62 (231,585.40, 306,644.39)	11.45 (9.83, 13.10)	−3.42 (−3.75, −3.09)
	Deaths	237,737.54 (169,947.96, 279,106.94)	28.46 (20.74, 33.37)	252,503.22 (216,053.93, 290,796.13)	10.86 (9.27, 12.42)	−3.59 (−3.93, −3.24)
	DALYs	6,399,999.56 (4,487,101.98, 7,492,387.26)	709.77 (503.35, 833.81)	5,697,007.31 (5,041,778.66, 6,514,368.89)	242.87 (215.07, 277.01)	−3.96 (−4.32, −3.60)
Germany	Incidence	4,586.69 (4,320.49, 4,869.73)	3.80 (3.59, 4.03)	9,115.16 (8,273.08, 10,003.98)	4.99 (4.56, 5.51)	0.57 (0.43, 0.72)
	Deaths	4,307.72 (4,094.10, 4,534.25)	3.50 (3.32, 3.68)	7,777.34 (7,151.79, 8,378.86)	4.06 (3.74, 4.38)	0.18 (0.09, 0.27)
	DALYs	112,351.75 (106,799.68, 117,948.22)	96.68 (91.79, 101.59)	168,162.12 (154,644.65, 181,492.84)	99.00 (90.59, 107.26)	−0.24 (−0.34, −0.13)
India	Incidence	17,732.74 (13,730.09, 23,243.43)	3.71 (2.89, 4.91)	97,320.91 (76,412.28, 122,701.69)	7.77 (6.09, 9.76)	2.21 (1.64, 2.78)
	Deaths	17,108.07 (13,329.83, 22,525.69)	3.76 (2.94, 4.98)	95,870.99 (75,277.71, 121,012.62)	7.95 (6.19, 9.95)	2.25 (1.68, 2.83)
	DALYs	514,135.98 (395,565.37, 665,985.10)	98.93 (76.66, 130.12)	2,587,553.59 (2,001,671.40, 3,269,787.73)	195.92 (151.70, 248.23)	2.02 (1.47, 2.58)
Japan	Incidence	11,391.34 (10,641.66, 12,177.71)	6.66 (6.20, 7.13)	22,611.83 (20,015.27, 24,387.21)	6.21 (5.61, 6.66)	−0.42 (−0.67, −0.17)
	Deaths	8,612.52 (8,079.42, 9,058.97)	5.08 (4.75, 5.35)	15,203.78 (13,361.65, 16,449.46)	3.86 (3.50, 4.15)	−1.04 (−1.18, −0.91)
	DALYs	206,842.33 (195,563.28, 216,341.11)	120.15 (113.48, 125.83)	271,603.10 (244,360.61, 291,513.42)	84.06 (77.25, 89.98)	−1.37 (−1.57, −1.18)
United States of America	Incidence	13,369.44 (12,136.75, 14,682.54)	4.32 (3.94, 4.74)	24,352.20 (21,505.93, 26,780.38)	4.15 (3.67, 4.55)	−0.23 (−0.40, −0.06)
	Deaths	11,776.47 (10,806.72, 12,666.96)	3.76 (3.46, 4.04)	21,128.68 (19,217.66, 23,095.29)	3.51 (3.19, 3.83)	−0.28 (−0.42, −0.14)
	DALYs	286,162.46 (265,015.46, 306,522.14)	96.08 (88.99, 103.15)	471,647.74 (430,084.62, 514,895.66)	84.02 (76.56, 91.25)	−0.49 (−0.62, −0.36)

CI, confidence interval; DALYs, disability-adjusted life years; EAPC, estimated annual percentage change; UI, uncertainty interval.

Age and gender differences

The disease burden of esophageal cancer shows significant differences by age and sex across five countries (China, Germany, India, Japan, and the United States). In 2023, the disease burden among men in China, Germany, Japan, and the United States was significantly higher than that among women, while in India, the disease burden among men was only slightly higher than that among women. The disease burden across all populations increases steadily with age. In terms of incidence, rates peaked in the 80–84 age group in China and Germany, while in India, Japan, and the United States, the peak occurred in the 85–89 age group. In terms of mortality, rates were highest in the 90–94 age group in China, India, and the United States, while in Germany and Japan, mortality rates rose with age. Regarding DALYs, China had the highest DALYs rate in the 75–79 age group, Germany in the 65–69 age group, India in the 85–89 age group, and Japan and the United States in the 70–74 age group (Figure 1). Overall, elderly males constitute the core high-risk population bearing the majority of esophageal cancer burden, with distinct epidemiological profiles across different national settings.

Figure 1.

The esophageal cancer patients of each age group in males and females with incidence, death, and DALYs number and their age-standardized rate (per 100,000 population) in 2023. (A) Incidence. (B) Mortality. (C) DALYs. DALYs, disability-adjusted life years; UI, uncertainty interval.

Time and age trend changes

From 1990 to 2023, the disease burden of esophageal cancer in these five countries exhibited distinct temporal trends. China bore the heaviest burden; however, from 1990 to 2023, its incidence, mortality, and DALYs rates declined substantially, indicating a progressive alleviation of the disease burden. Notably, India experienced increasing trends in incidence, mortality, and DALYs rates over the same period, suggesting a gradually worsening burden. In Germany, the incidence of esophageal cancer rose slowly, while mortality and DALYs rates remained stable. The disease burden in the United States stayed relatively stable. In Japan, incidence rates remained stable, whereas mortality and DALYs rates decreased slightly. Furthermore, we observed that in all five countries, the burden of esophageal cancer was consistently higher in males than in females (Figure 2).

Figure 2.

Age-standardized incidence (A), mortality (B), and DALYs (C) rates of esophageal cancer in five countries (1990–2023), stratified by sex. DALYs, disability-adjusted life years.

The disease burden of esophageal cancer also showed distinct patterns across different age groups. Since 1990, China and India have exhibited opposite trends: in China, the incidence, mortality, and DALYs rates generally declined across all age groups, particularly in the 70–74 age group, although the absolute rates remained the highest among the five countries; in India, the burden increased markedly, with the growth concentrated in the elderly, especially those aged 85 years and above. In Germany, the burden among individuals aged over 60 years increased over time. In Japan, the burden across age groups remained relatively stable, with notable declines in DALY rates observed in the 55–59 and 60–64 age groups. In the United States, incidence and mortality rates increased among those aged over 75 years, while DALYs rates showed no significant change. Moreover, in all five countries, the elderly bore the heaviest disease burden (Figure 3). These disparities suggest that the pattern of esophageal cancer burden is closely associated with the stage of economic development in each country.

Figure 3.

The esophageal cancer patients of different age groups with age-standardized incidence (A), mortality (B), and DALYs (C) rates from 1990 to 2023. DALYs, disability-adjusted life years.

Joinpoint regression analysis of disease burden

From 1990 to 2023, trends in the disease burden varied across the five countries. In China, ASIR, ASMR, and ASDR generally showed a downward trend, but rebounded after 2019 or 2020. In Germany, ASIR showed an overall upward trend, but declined between 2007 and 2018. ASMR rose from 1990 to 1996, followed by fluctuations, and rose again in 2011. ASDR peaked in 1996 and then declined in a wave-like pattern, reaching its lowest point in 2016. In India, the three disease burden indicators fluctuated slightly before 2012 and showed a consistent upward trend after 2012. In Japan, the ASIR rose from 1990 to 2008, then declined significantly, before rebounding after 2021. Similar to the ASIR, the ASMR and ASDR rebounded after declining significantly between 2001 and 2021. In the United States, the disease burden indicators first rose and then fell, peaking in 2001 or 2002 (Figure 4).

Figure 4.

Joinpoint regression analysis of the age-standardized incidence (A), mortality (B), and DALYs (C) rates for esophageal cancer patients from 1990 to 2023. APC, annual percentage change; DALYs, disability-adjusted life years.

APC analysis results

To delineate the independent influences of age, time period, and birth cohort on disease risk, an APC analysis was conducted in five countries. In China, the age effects of the three indicators first increased and then decreased with advancing age, following an “inverted U-shaped” curve. Both period and cohort effects showed declining trends. The net drift analysis indicated that the estimated annual percentage changes were negative across all age groups, with slower declines in the elderly population. In Germany, the age effects of ASIR and ASMR gradually increased with age, while the age effect of ASDR peaked in the elderly and then declined in a wave-like manner. The period effect for ASIR showed a wave-like upward trend, whereas those for ASMR and ASDR peaked during 1994–1998, then declined, and rebounded after 2009–2013. The cohort effects for all three indicators first increased and then decreased. The net drift showed an upward trend, and local drift analysis further indicated declining trends in younger and middle-aged groups, but increasing trends in the elderly. In India, the age, period, and cohort effects of disease burden all exhibited upward trends. The net drift followed a “U-shaped” pattern, with the lowest increases observed in middle-aged and elderly groups. For Japan and the United States, the age effects were similar: the age effects for incidence and mortality were highest in the elderly, whereas the age effect for DALYs declined in the very elderly (aged 80 years and above). Regarding period effects, Japan showed an overall declining trend, while the United States exhibited an “inverted U-shaped” pattern with a decline after 2003. Cohort effects indicated fluctuations in disease risk in both countries. Local drift analysis revealed declining trends in younger and middle-aged groups, but increasing trends in the elderly in both countries (Figure 5; Figures S1,S2).

Figure 5.

Age, period and cohort effects on esophageal cancer incidence in five countries. (A) China. (B) Germany. (C) India. (D) Japan. (E) United States of America.

Risk factor-attributable burden characteristics

Based on the attributable risk analysis for esophageal cancer, the deaths and DALYs attributed to the disease in the five countries were primarily driven by smoking, high alcohol use, low vegetable intake, and chewing tobacco. In China, Germany, Japan, and the United States, smoking was the leading contributor to deaths and DALYs, followed by high alcohol use, with chewing tobacco contributing the least. Compared with China, high alcohol use and low vegetable intake played more prominent roles in deaths and DALYs in Germany. In Japan and the United States, the proportional contribution of low vegetable intake was also higher than that in China. Notably, India exhibited a pattern co-driven by low vegetable intake and chewing tobacco, with both contributing similar proportions, followed by smoking, while high alcohol use made the smallest contribution (Figure 6A,6B; Figure S3).

Figure 6.

Disease burden of esophageal cancer attributable to risk factors in five countries. (A) Cross-country comparison of mortality proportions attributable to risk factors. (B) Cross-country comparison of DALYs proportions attributable to risk factors. (C) Temporal trends of mortality proportions attributable to risk factors. (D) Temporal trends of DALYs proportions attributable to risk factors. DALYs, disability-adjusted life years.

From 1990 to 2023, the contributions of the four risk factors to deaths and DALYs varied over time across the five countries. In China, the proportion of deaths and DALYs attributed to smoking rose substantially, whereas it declined in the other countries. The proportion attributable to high alcohol use increased in China and the United States, remained stable in Germany, and decreased in India and Japan. The proportion due to low vegetable intake decreased in China and India, remained stable in Germany and the United States, and first declined and then increased over time in Japan. Except for India, the proportion of deaths and DALYs caused by chewing tobacco remained the lowest and stable in the other populations, whereas in India it declined over time (Figure 6C,6D).

Decomposition analysis of disease burden

Decomposition analysis of changes in the esophageal cancer burden from 1990 to 2023 across the five countries revealed marked differences in driving factors. In China, epidemiological changes (shifts in disease patterns) were the primary factor reducing the incidence of esophageal cancer (−267,524.84, −1,105.34%), while aging effects (152,270.27, 629.14%) and population changes (139,457.44, 576.2%) contributed to an overall increase in incident cases. In Germany, aging led to a decrease of 901.47 cases (−19.86%), whereas population changes and epidemiological changes had positive effects, increasing cases by 3,232.26 (71.21%) and 2,208.05 (48.65%), respectively. In India, aging, population changes, and epidemiological changes all contributed to the rise in esophageal cancer incidence, with increases of 24,220.39 (30.43%), 23,225.20 (29.18%), and 32,142.58 (40.39%) cases, respectively. Japan and the United States shared a similar pattern: aging and population changes had positive effects on incidence, while epidemiological changes had the opposite effect, with population changes being the dominant factor influencing incidence (Figure 7A, Table 2). Similar patterns were observed for deaths and DALYs, where epidemiological changes negatively impacted the disease burden in China, Japan, and the United States, and were the main factor alleviating the burden in China. In Germany, population changes were the principal factor affecting the burden, whereas in India, epidemiological changes played the leading role. Moreover, population changes consistently increased the disease burden across all five countries (Figure 7B,7C).

Figure 7.

Decomposition analysis of esophageal cancer burden in five countries: (A) Incidence. (B) Mortality. (C) DALYs. Positive x-axis values indicate increasing metrics; negative values indicate decreasing metrics. Black dots represent the total impact of three factors. DALYs, disability-adjusted life years.

Table 2. Decomposition of changes in esophageal cancer burden in five countries from 1990 to 2023.

Location	Measure	Overall difference	Aging	Population	Epidemiological change	Aging percentage (%)	Population percentage (%)	Epidemiological change percentage (%)
China	Incidence	24,202.87	152,270.27	139,457.44	−267,524.84	629.14	576.2	−1,105.34
	Deaths	14,765.69	157,006.97	135,699.96	−277,941.24	1,063.32	919.02	−1,882.34
	DALYs	−702,992.19	3,334,332.93	3,403,133.11	−7,440,458.24	474.31	484.09	−1,058.4
Germany	Incidence	4,538.84	−901.47	3,232.26	2,208.05	−19.86	71.21	48.65
	Deaths	3,491.67	−604.91	2,876.87	1,219.71	−17.32	82.39	34.93
	DALYs	55,880.86	−25,728.4	67,945.67	13,663.58	−46.04	121.59	24.45
India	Incidence	79,588.17	24,220.39	23,225.2	32,142.58	30.43	29.18	40.39
	Deaths	78,762.92	24,972.23	22,687.4	31,103.3	31.71	28.8	39.49
	DALYs	2,073,417.61	597,270.13	635,905.88	840,241.59	28.81	30.67	40.52
Japan	Incidence	11,220.39	3,738.32	7,920.54	−438.47	33.32	70.59	−3.91
	Deaths	6,591.18	3,572.56	5,675.51	−2,656.89	54.2	86.11	−40.31
	DALYs	64,760.01	20,376.57	117,104.51	−72,721.07	31.46	180.83	−112.29
United States of America	Incidence	10,979.44	2,353.67	8,992.77	−367	21.44	81.91	−3.34
	Deaths	9,347.83	2,238.6	7,859.32	−750.08	23.95	84.08	−8.02
	DALYs	185,448.21	44,346.24	183,110.11	−42,008.14	23.91	98.74	−22.65

DALYs, disability-adjusted life years.

BAPC prediction analysis

The BAPC model was used to project the disease burden of esophageal cancer in the five countries from 2024 to 2050. According to the model projections, the disease burden is projected to show a steady downward trend in China, Germany, and Japan, an upward trend in India, and a decline followed by an increase in the United States over this period. The projected ASIR (per 100,000 population) in 2050 is 7.41 (95% UI: 3.63, 11.19) for China, 6.58 (95% UI: 5.22, 7.92) for Germany, 27.67 (95% UI: 10.36, 44.97) for India, 7.24 (95% UI: 5.47, 9.01) for Japan, and 7.03 (95% UI: 5.77, 8.28) for the United States (Figure 8A). The trends for ASMR and ASDR are similar to those for ASIR, with the ASMR and ASDR for China projected to decrease to 5.7 (95% UI: 2.69, 8.71) and 120.69 (95% UI: 58.22, 183.15), respectively; for Germany to 4.4 (95% UI: 3.37, 5.42) and 97.08 (95% UI: 76.54, 117.61); for Japan to 3.45 (95% UI: 2.58, 4.31) and 81.64 (95% UI: 64.95, 98.32); for India to 26.23 (95% UI: 9.46, 42.99) and 708.84 (95% UI: 245.74, 1,171.92); and for the United States to 5.43 (95% UI: 4.57, 6.28) and 137.29 (95% UI: 108.62, 165.92) (Figure 8B,8C). Notably, although the incidence, mortality, and DALYs rates in the United States are projected to decline first and then increase, the absolute number of cases is expected to show a steady upward trend (Figure S4). In the out-of-sample validation (training period 1990–2013; validation period 2014–2023), the BAPC model showed good overall forecast accuracy (Figure S5). Across the five countries, all MAPEs were below 15%: rate estimates ranged from 0.77% (India, incidence) to 13.54% (Germany, DALYs rate), and absolute numbers ranged from 0.71% (India, deaths) to 9.58% (China, deaths). India had the highest agreement (all MAPEs <1.1%), whereas China had the largest errors (rate MAPEs 5.88–12.22%), mainly reflecting underestimation, suggesting that extrapolation errors may accumulate over longer horizons in rapidly transitioning settings.

Figure 8.

Temporal trends of age-standardized rates of esophageal cancer in five countries from 1990 to 2023, with projections for 2024–2050. (A) Incidence. (B) Mortality. (C) DALYs. ASR, age-standardized rate; DALYs, disability-adjusted life years.

Discussion

In our study, leveraging the 2023 GBD database we performed, for the first time, a systematic analysis of the esophageal cancer burden in the world’s five largest economies-China, Germany, India, Japan, and the United States-from 1990 to 2023. Our findings reveal substantial heterogeneity among these countries in terms of disease trajectories, risk profiles, and future projections. The scientific value of this comparative analysis lies in the marked heterogeneity among these countries: they span distinct socioeconomic development stages (from high-income to lower-middle-income), exhibit divergent demographic structures (from severely aged to predominantly young populations), and present contrasting esophageal cancer subtype profiles (squamous-cell-predominant in Asian countries versus adenocarcinoma-predominant in Western nations). Such diversity allows us, within a unified GBD analytical framework, to disentangle the relative contributions of demographic transitions, risk factor exposures (smoking, alcohol, chewing tobacco, and dietary patterns), and healthcare system capacities to the divergent disease trajectories observed across these economically influential nations. Moreover, insights derived from this comparison may inform tailored cancer control strategies not only for these countries but also for other nations at similar developmental stages. In our findings, although China maintained the highest absolute burden, its ASIR declined markedly, reflecting a successful epidemiological transition. In contrast, India experienced a persistent increase in both incidence and mortality. The remaining three high-income countries exhibited relatively stable and lower burdens, yet with distinct national patterns. This heterogeneous pattern underscores the complex roles of population aging, socioeconomic development, risk factor exposure, and healthcare system capacity in shaping the epidemiological landscape of esophageal cancer among the world’s most economically influential countries.

The disparate esophageal cancer burdens across the five countries can be attributed to their different stages of socioeconomic development and corresponding public health responses. In China, the age-standardized incidence and mortality rates declined substantially-from 28.23 to 11.45 per 100,000 for incidence and from 28.46 to 10.86 per 100,000 for mortality-representing the largest decrease among the five countries. Joinpoint regression analysis indicated that the most pronounced decline occurred during 2007–2013, an achievement likely attributable to a series of prevention and control policies implemented by the Chinese government, including the issuance of the Outline of the Chinese Cancer Prevention and Control Plan (2004–2010) and the inclusion of esophageal cancer in the priority screening program for high-incidence tumors in 2012 (26,27). Moreover, advances in minimally invasive surgical techniques and the application of immune checkpoint inhibitors have improved patient survival and quality of life (28,29). However, Joinpoint regression also revealed a significant rebound in incidence after 2019, possibly related to enhanced cancer control measures in China following 2019, including expanded screening coverage and intensified early detection efforts, which may have led to increased case detection (30). In India, a concerning trend emerged: the ASIR more than doubled from 3.71 to 7.77 per 100,000, with EAPCs of 2.21% for incidence and 2.25% for mortality. This escalating burden reflects the ongoing epidemiological transition in the country, potentially driven by changing dietary habits, chewing tobacco and betel nut, and insufficient healthcare service provision (31). The United States, Japan, and Germany, as high-income countries with well-established healthcare systems, more effective preventive measures, and earlier detection, maintained relatively low and stable incidence and mortality rates. However, Japan exhibited higher incidence rates compared with the United States and Germany, possibly reflecting higher prevalence of alcohol consumption and smoking, as well as unhealthy dietary habits such as low fruit and vegetable intake, among its male population (32). These intercountry disparities highlight that economic prosperity alone does not guarantee a low disease burden; rather, sustained investment in primary prevention and early detection is essential.

APC and decomposition analyses further elucidate the mechanisms driving these heterogeneous trends. The age effects indicated that elderly males constitute the core high-risk group in all five countries, corroborating previous findings (33). Notably, China exhibited an “inverted U-shaped” age-effect curve, with incidence peaking at age 70–74 years, whereas Germany, the United States, and Japan showed peak incidence at older ages (85–89 years), possibly related to the higher life expectancy in developed countries. The cohort effects revealed compelling generational shifts: in China, disease risk progressively declined in more recent birth cohorts, which may suggest that government policies on early cancer prevention and control have contributed to this favorable trend. Conversely, India showed increasing risk in recent cohorts, foreshadowing an unfavorable future trajectory unless effective interventions are implemented. Decomposition analysis revealed that while population aging and population growth exerted substantial upward pressure, favorable epidemiological changes (declining age-specific incidence rates) strongly mitigated the disease burden in China. This finding carries profound implications: current prevention programs can, to some extent, offset the burden imposed by population aging and growth, possibly representing a phase of favorable epidemiological transition. In stark contrast, in India, aging, population growth, and unfavorable epidemiological changes jointly contributed to the rising burden, potentially reflecting deficiencies in screening and early diagnosis (34).

Our risk factor attribution analysis identified country-specific priorities that are critical for precision public health planning. In China, Germany, Japan, and the United States, smoking emerged as the predominant driver of esophageal cancer deaths and DALYs, supporting the continued emphasis on tobacco control. India exhibited a distinctive dual-driver pattern, where chewing tobacco and low vegetable intake jointly dominated the disease burden, whereas the contributions of smoking and high alcohol use were comparatively lower, possibly reflecting the distinct risk exposure profile of the country (35,36). Furthermore, the temporal trends in risk factor contributions were equally instructive: during the study period, the proportion of deaths and DALYs attributable to smoking in China increased substantially, possibly related to the country’s relatively weak tobacco control measures and deeply entrenched smoking culture (37). Therefore, strengthening tobacco regulation and enhancing public awareness of the health hazards posed by tobacco are of paramount importance.

The BAPC projections to 2050 depict the future landscape of esophageal cancer burden. In China, both the incidence rate and the absolute number of incident cases are projected to continue declining. The projected trajectory for India is concerning, with the incidence rate reaching 27.67 per 100,000 by 2050, heralding enormous pressure on esophageal cancer prevention and treatment. The United States exhibits a V-shaped pattern, with the ASIR projected to decline first and then rebound, possibly driven by an increase in esophageal adenocarcinoma related to the obesity epidemic and gastroesophageal reflux disease-factors not fully captured by traditional risk factor models (38). These projections may help inform context-specific policy considerations. For China, maintaining and expanding endoscopic screening coverage while continuing tobacco cessation and alcohol harm reduction efforts could be explored to sustain the downward trend (39). For India, integrating chewing tobacco cessation into primary health care, developing population-based screening infrastructure, and promoting healthy lifestyles may be worth considering. For the United States, Japan, and Germany, sustaining existing gains, reviewing screening strategies in the elderly, and investigating emerging risk factors such as obesity and Barrett’s esophagus could be considered. Moreover, all five countries may face a rise in absolute case numbers due to population aging, which could require strengthening healthcare system capacity, workforce training, and resource allocation to manage the growing number of elderly patients with oesophageal cancer requiring multimodal therapy. In addition, we validated the model and found that the projected trajectories closely matched the observed values, with most observations falling within the 95% UIs, supporting the credibility of projections to 2050; however, long-term estimates should be interpreted cautiously because the UIs widen over time.

Our study has several limitations. First, as with all analyses based on the GBD study, our findings depend on the quality and completeness of the original data sources. However, epidemiological surveys in many middle and low income countries are incomplete, which may introduce bias and reduce the accuracy of burden estimates. Second, the GBD database only includes risk factors such as low vegetable intake, smoking, alcohol use, and chewing tobacco; other potential factors, such as nitrosamine intake, consumption of overheated beverages, and genetic mutations, were not incorporated in the current analysis (33). Third, the BAPC projections of disease burden are based on extrapolations of historical trends, but future breakthroughs in prevention, screening technologies, or treatment, as well as unforeseeable public health emergencies or policy shifts, may affect the accuracy of the predictions. Finally, the GBD database does not distinguish between subtypes of esophageal cancer (e.g., esophageal squamous cell carcinoma vs. esophageal adenocarcinoma), despite their distinct epidemiological features and etiological pathways. This may obscure important etiological heterogeneity, particularly when comparing Asian countries (where esophageal squamous cell carcinoma predominates and is largely driven by smoking, alcohol, and dietary factors) with Western countries (where esophageal adenocarcinoma has been rising, largely in association with obesity, gastroesophageal reflux disease, and Barrett’s esophagus).The lack of subtype-specific analyses limits the precision of our study and may obscure heterogeneous trends across different populations (40).

Conclusions

Based on GBD 2023 data, this study systematically compared the esophageal cancer burden from 1990 to 2023 and projected trends to 2050 across the world’s five largest economies. Our results showed that China experienced a marked decline in ASRs, yet its absolute burden remained the highest, with population aging and growth offsetting the epidemiological improvements. In contrast, India faces escalating pressure from rising incidence and mortality. The United States, Japan, and Germany maintained relatively low levels, albeit with distinct national characteristics. Smoking was the predominant risk factor in most countries, whereas India was primarily driven by chewing tobacco and low vegetable intake. BAPC projections indicated a surge in disease burden in India, a potential rebound in the United States, and continued declines in China, Germany, and Japan. These findings may help inform tailored cancer control strategies aligned with each country’s demographic structure and risk factor profile, such as strengthening tobacco and alcohol control, promoting healthier diets, and considering screening in high-incidence regions.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.