Introduction
Bladder cancer (BC) is one of the most common malignancies of the urinary system worldwide. According to GLOBOCAN 2022, BC ranked ninth for cancer incidence and thirteenth for cancer mortality globally, with an estimated 614,298 new cases and 220,596 deaths in 2022, indicating a substantial global health burden.1 BC is also characterized by a high risk of recurrence, with reported recurrence rates ranging from 50% to 70%; therefore, many patients require prolonged, and sometimes lifelong, surveillance because of persistent risks of recurrence and progression.2 Previous studies have further shown that BC has one of the highest lifetime treatment costs per patient among all cancers, placing considerable demands on health-care systems and substantial economic and psychological burdens on patients.3
China bears a particularly large burden of BC, with the highest numbers of incident cases and deaths worldwide, accounting for approximately 15% of global incident cases and 19% of global deaths.4 BC occurs predominantly in older adults,5 and several modifiable risk factors, including tobacco use and metabolic conditions such as diabetes mellitus, contribute to its development and progression.6,7 In the context of population aging and changing patterns of modifiable risk exposure,8,9 a clearer understanding of the current burden, temporal trends, and potential drivers of BC is essential for guiding prevention and control strategies. Therefore, using data from the Global Burden of Disease (GBD) Study 2023, this study systematically assessed the burden of BC in China by characterizing its current burden profile, temporal trends, age-period-cohort patterns, and future projections. Incidence, mortality, and disability-adjusted life years (DALYs) were used as the main burden indicators. Because population-based organized screening for BC is not currently available in China, and targeted prevention and early-detection strategies are increasingly needed, this study also aimed to identify high-risk age groups and quantify the contributions of modifiable risk factors to inform future risk-stratified screening, early detection, and primary prevention strategies.
Results
Profiles of BC burden in China, 2023
Table 1 presents the burden profile of bladder cancer in China in 2023 for the overall population and by sex. An estimated 105,632 incident cases and 44,728 deaths from bladder cancer occurred in China in 2023, corresponding to a crude incidence rate of 7.38 per 100,000 and a crude mortality rate of 3.13 per 100,000. The age-standardized incidence rate (ASIR) and age-standardized mortality rate (ASMR) were 4.66 and 2.00 per 100,000, respectively. Bladder cancer accounted for 902,946 DALYs, with a crude DALY rate of 63.11 per 100,000 and an age-standardized DALY rate (ASDR) of 39.67 per 100,000. A marked sex disparity was observed across all burden indicators. Males accounted for most incident cases, deaths, and DALYs, with 85,985 new cases, 34,936 deaths, and 719,427 DALYs, compared with 19,647 cases, 9,793 deaths, and 183,520 DALYs in females. Age-standardized rates were consistently higher in males than in females: the ASIR was nearly five times higher (7.98 vs. 1.66 per 100,000), the ASMR was more than four times higher (3.42 vs. 0.81 per 100,000), and the ASDR was also substantially higher (66.57 vs. 15.45 per 100,000). Among modifiable risk factors, tobacco exposure was the leading contributor to bladder cancer deaths in China. Overall, 38.06% of bladder cancer deaths were attributable to smoking, with a substantially higher proportion in males than in females (46.60% vs. 7.62%). By contrast, the proportion of deaths attributable to high fasting plasma glucose (FPG) was lower overall (10.37%) and showed only a small sex difference, accounting for 10.42% of deaths in males and 10.22% in females.
Table 1Burden profile of bladder cancer in China, 2023
| Level | Incidence
| Mortality
| DALY
| Risk factor
|
|---|
| Number | CIR | ASIR | Number | CMR | ASMR | Number | CDR | ASDR | Percentage of deaths attributable to tobacco | Percentage of deaths attributable to high FPG |
|---|
| Overall | 105,632 (85,504, 126,385) | 7.38 (5.98, 8.83) | 4.66 (3.78, 5.54) | 44,728 (38,358, 50,790) | 3.13 (2.68, 3.55) | 2.00 (1.71, 2.28) | 902,946 (783,966, 1,020,874) | 63.11 (54.80, 71.36) | 39.67 (34.45, 44.92) | 38.06% (32.29%, 43.47%) | 10.37% (6.30%, 15.55%) |
| Male | 85,985 (69,594, 104,250) | 11.74 (9.50, 14.24) | 7.98 (6.54, 9.63) | 34,936 (30,066, 40,863) | 4.77 (4.11, 5.58) | 3.42 (2.91, 4.03) | 719,427 (621,521, 840,770) | 98.24 (84.87, 114.81) | 66.57 (57.39, 76.71) | 46.60% (40.00%, 52.54%) | 10.42% (6.36%, 15.74%) |
| Female | 19,647 (14,824, 26,549) | 2.81 (2.12, 3.80) | 1.66 (1.26, 2.26) | 9,793 (7,680, 13,341) | 1.40 (1.10, 1.91) | 0.81 (0.64, 1.11) | 183,520 (147,188, 240,695) | 26.28 (21.08, 34.47) | 15.45 (12.39, 20.21) | 7.62% (5.33%, 10.78%) | 10.22% (6.13%, 15.25%) |
Temporal trends of BC burden in China, 1990 to 2023
Figure 1 and Supplementary Table 1 present temporal trends in the incidence, mortality, and DALY burden of bladder cancer in China from 1990 to 2023. The crude incidence rate showed a significant overall upward trend, with an average annual increase of 2.18% (95% CI: 2.05% to 2.32%, P < 0.001). In contrast, the ASIR showed a slight but significant overall decline during the same period, with an AAPC of −0.32% (95% CI: −0.44% to −0.17%, P < 0.001). However, marked period-specific fluctuations were observed. The ASIR declined continuously from 1990 to 2014, with the steepest decrease during 2000–2003 (APC = −3.97%, 95% CI: −4.80% to −1.76%, P = 0.03). After 2014, the trend reversed and began to increase, with a particularly pronounced rise during 2020–2023 (APC = 5.05%, 95% CI: 3.41% to 8.27%, P < 0.001). Sex-specific analyses showed that the increase in crude incidence was more pronounced in males than in females, with AAPCs of 2.29% (95% CI: 2.13% to 2.47%) and 0.86% (95% CI: 0.77% to 0.95%), respectively. The overall decline in ASIR was more marked in females than in males, with AAPCs of −1.56% (95% CI: −1.65% to −1.47%) and −0.32% (95% CI: −0.49% to −0.14%), respectively.
For mortality, the crude mortality rate also showed a significant overall increase from 1990 to 2023, with an average annual increase of 1.52% (95% CI: 1.39% to 1.70%, P < 0.001). During the same period, the ASMR showed an overall decline, with an AAPC of −1.31% (95% CI: −1.44% to −1.15%, P < 0.001), although the decline was not monotonic. Specifically, the ASMR increased during 2014–2017 (APC = 3.46%, 95% CI: 1.81% to 4.47%), declined again during 2017–2020, and then increased during 2020–2023 (APC = 4.39%, 95% CI: 2.77% to 7.55%). In males and females, the AAPCs of the crude mortality rate were 0.50% and 1.76%, respectively, whereas the corresponding AAPCs of the ASMR were −1.09% and −2.31%, respectively (Fig. 1b, Supplementary Tables 1 and 2). Given the high recurrence rate of bladder cancer and its substantial effect on quality of life, temporal trends in DALYs were also analyzed. The crude DALY rate increased overall from 1990 to 2023 (AAPC = 0.79%, 95% CI: 0.67% to 0.96%, P < 0.001), whereas a decline was observed in females (AAPC = −0.35%, 95% CI: −0.43% to −0.27%, P < 0.001). Consistent with the incidence and mortality patterns, the ASDR declined overall from 1990 to 2023 (AAPC = −1.62%, 95% CI: −1.76% to −1.45%, P < 0.001), but rebounded in recent years, increasing during 2020–2023 (APC = 4.25%, 95% CI: 2.63% to 7.39%, P < 0.001) (Fig. 1c, Supplementary Tables 1 and 2).
Trends in bladder cancer burden attributable to modifiable risk factors were also assessed. For bladder cancer deaths attributable to tobacco, the ASMR showed an overall decline (AAPC = −1.53%, 95% CI: −1.71% to −1.33%), although an upward, non-significant trend was observed during 2013–2023 (APC = 0.33%, P = 0.33). For bladder cancer deaths attributable to high FPG, the ASMR also declined overall (AAPC = −1.02%, 95% CI: −1.13% to −0.89%). A similar pattern of an initial decrease followed by a subsequent increase was observed (Supplementary Tables 1 and 2, Supplementary Fig. 1). Diagnostic statistics for the main Joinpoint models are presented in Supplementary Table 3. Across burden indicators and sex groups, Bayesian information criterion values ranged from −5.26 to −1.44 and root mean squared error values ranged from 0.05 to 0.38. In sensitivity analyses varying the maximum number of joinpoints, the results were largely consistent with the main analysis. Reducing the maximum number from six to five did not change the selected joinpoints, AAPCs, recent trend segments, or corresponding APCs. When the maximum number was further reduced to four, fewer joinpoints were selected for some indicators, but the directions of the AAPCs and recent APCs generally remained unchanged; the exception was that the recent APC for female ASMR became statistically non-significant (Supplementary Table 4). In the sensitivity analysis excluding data from 2020 to 2023, AAPC estimates were lower across all indicators, while their directions remained largely unchanged, except for the crude mortality rate in females (Supplementary Table 5).
Age, period, and cohort effect analysis of BC burden in China, 1994 to 2023
In the age-period-cohort modelling, age and period were categorized into equal five-year intervals to ensure accurate derivation of birth cohorts, enhance model identifiability, and improve the stability and interpretability of age, period, and cohort effect estimates.21 Accordingly, the analysis used data from 1994 to 2023, allowing complete and comparable five-year period groupings across the study period. The age-specific temporal patterns of BC burden in China from 1994 to 2023 are shown in Figure 2 using local drifts and longitudinal age curves. For both incidence and mortality, local drift values were below zero in most age groups, indicating overall declines in age-specific rates over time. However, local drift increased progressively with age and approached or exceeded zero in the oldest age groups. For incidence, local drift values were 0.35% (95% CI: −0.07%, 0.78%) in the 85–89-year age group and 0.64% (95% CI: −0.31%, 1.61%) in the 90–94-year age group. For mortality, the corresponding values were −0.70% (95% CI: −0.97%, −0.43%) and −0.18% (95% CI: −0.72%, 0.36%), respectively, indicating stabilization and possible reversal of age-specific trends in the oldest populations. Across most age groups, females had lower local drift values than males, indicating a greater decline over time. The longitudinal age curves showed marked age effects for both incidence and mortality. Age-specific incidence and mortality rates remained low at younger ages and increased substantially with advancing age, particularly after 60 years. Males consistently had higher rates than females across nearly all age groups.
Figure 3 presents the period and cohort RRs for bladder cancer incidence and mortality in China by sex. For both incidence and mortality, period RRs generally decreased over time, with earlier periods showing RRs above 1.0 and later periods showing RRs below 1.0 relative to the reference period (2004–2008). This decline was more pronounced in females. In males and in the overall population, however, incidence period effects rebounded from 2014–2018 onward, and the decline in mortality period effects became less pronounced over the same period. Cohort RRs for both incidence and mortality declined progressively across successive birth cohorts. For incidence, compared with the reference cohort (1951–1956), earlier birth cohorts had elevated risks, with RRs of 1.13 for both sexes, 1.09 for males, and 2.17 for females in the 1901–1906 cohort. Reduced risks were observed in later cohorts, with RRs decreasing to 0.37 for both sexes, 0.44 for males, and 0.21 for females in the 2001–2006 cohort. A similar pattern was observed for mortality. The downward cohort trend was particularly pronounced among females, suggesting substantial generational improvements in bladder cancer risk. Model diagnostic statistics are presented in Supplementary Table 6. No age-period cell had a dispersion-adjusted APC residual greater than 3, and the observed-versus-fitted rate mean absolute percentage error (MAPE) ranged from 2.97% to 3.47% across the six age-period-cohort models. In the sensitivity analysis using 10-year age groups and 10-year periods, the overall directions and patterns of age, period, and cohort effects were generally consistent with those in the main analysis (Supplementary Figs. 2 and 3).
Projections of BC burden in China, 2024 to 2030
Short-term projections of bladder cancer incidence and mortality in China from 2024 to 2030 were generated using BAPC models. The crude incidence rate was projected to increase steadily for both sexes combined, from 8.5 per 100,000 population in 2024 to 10.9 per 100,000 population in 2030 (Fig. 4a). Similar increases were projected by sex, from 12.6 to 16.1 per 100,000 population in males and from 3.0 to 4.4 per 100,000 population in females. The crude mortality rate was also projected to increase during the same period (Fig. 4b). For both sexes combined, the crude mortality rate was projected to rise from 3.3 per 100,000 population in 2024 to 4.0 per 100,000 population in 2030. Sex-specific projections showed similar upward trends, from 4.5 to 5.7 per 100,000 population in males and from 1.9 to 2.2 per 100,000 population in females. For ASIRs, a slight increase was projected from 2024 to 2030 (Fig. 4c). The ASIR for both sexes combined was projected to increase marginally from 5.1 to 5.2 per 100,000 population. In sex-specific analyses, the ASIR was projected to increase from 8.3 to 8.5 per 100,000 population in males and from 1.7 to 2.0 per 100,000 population in females. ASMRs showed divergent trends by sex (Fig. 4d). The ASMR for both sexes combined was projected to decline from 2.0 per 100,000 population in 2024 to 1.8 per 100,000 population in 2030. A similar decline was projected in males, from 3.4 to 2.9 per 100,000 population, whereas the ASMR in females was projected to increase slightly from 0.8 to 0.9 per 100,000 population. Model fit and residual diagnostics for the six BAPC models are presented in Supplementary Table 7. Age-standardized rate MAPE values ranged from 0.13% to 2.76%, and the proportion of absolute Pearson residuals greater than 3 was 0% for five models and 2.35% for the incidence model in both sexes. In the sensitivity analysis using stronger smoothing priors, the projected trends were generally consistent with those from the main model (Supplementary Figs. 4 and 5).
Discussion
This study comprehensively profiled the burden of bladder cancer in China using GBD 2023 data, integrating current burden estimates, long-term temporal trends, age-period-cohort patterns, and future projections, with incidence, mortality, and DALYs as the key outcome measures. The findings provide an updated overview of the epidemiology and evolving burden of bladder cancer in China and offer evidence to support priority setting, policy development, and strengthening of cancer control strategies.
The temporal trend analyses showed a divergence between crude and age-standardized trends. Age-standardized rates for incidence, mortality, and DALYs (AAPC = −0.32%, −1.31%, and −1.62%, respectively) declined overall, which may reflect population-level improvements in risk factor control and clinical management in recent decades. Smoking is one of the most established risk factors for bladder cancer; in China, 38.06% of bladder cancer deaths in 2023 were attributable to smoking, and current smokers have been reported to have an approximately threefold higher risk of bladder cancer than never-smokers.6,22 Previous studies have shown that smoking prevalence and smoking intensity among Chinese adults declined between 2010 and 2018, indicating progress in tobacco control.23 This pattern is broadly consistent with our attributable-risk analysis, in which the ASMR attributable to tobacco declined overall, although this consistency should be interpreted as a population-level observation rather than direct evidence of causality. Long-term improvements in bladder cancer diagnosis and treatment may also have contributed to the declines in age-standardized mortality and DALY rates. Advances have included increasing standardization of transurethral resection, risk-adapted intravesical therapy, radical cystectomy, minimally invasive surgery, perioperative systemic treatment, and, more recently, immunotherapy and other novel systemic agents, all of which have expanded the treatment landscape for bladder cancer.24–26 However, crude incidence and mortality rates increased significantly (AAPC = 2.18% and 1.52%, respectively), suggesting that gains in risk control and clinical management may have been offset by population aging and expansion of high-risk age groups. Given the strongly age-related nature of bladder cancer, demographic aging has likely contributed substantially to the increasing absolute burden of the disease.22,27 These findings suggest that bladder cancer may continue to place increasing pressure on public health and health-care systems in China, underscoring the need for prevention and control strategies tailored to an aging population.
Although age-standardized incidence, mortality, and DALY rates generally declined from 1990 to 2023, these favorable trends reversed in recent years. Age-standardized incidence, mortality, and DALY rates began to rise after 2014, with more rapid increases during 2020–2023, as reflected by APCs of 5.05%, 4.39%, and 4.25%, respectively. The age-period-cohort analysis also supported this pattern: period and cohort RRs generally declined over time, suggesting long-term improvements in population-level conditions and risk profiles, but incidence period effects rebounded from 2014–2018 onward, and the decline in mortality period effects became less pronounced over the same period. These changes may be explained from several perspectives. First, improved case ascertainment and surveillance may have contributed to the observed increase. Over the past decade, China’s cancer registration and surveillance systems have continued to expand and improve, which may have enabled more complete capture of incident cases and deaths.28,29 Second, wider availability of diagnostic procedures, increased health-care utilization, and improved access to urological services may have increased the likelihood of bladder cancer diagnosis, particularly among older adults and populations with previously limited access to specialized care.30,31 Third, changes in the spectrum of risk exposures during rapid modernization in China may also have contributed to the recent reversal in bladder cancer burden. Previous studies have reported substantial transitions in dietary and lifestyle patterns, including increased consumption of sugar-sweetened beverages and highly processed foods, more sedentary behavior, and a rising burden of metabolic abnormalities such as obesity and diabetes.9,32–34 This interpretation is also consistent with our finding that the ASMR attributable to high FPG showed a recent upward trend. Finally, the increase in bladder cancer incidence and mortality during 2020–2023 should be interpreted cautiously. During the COVID-19 pandemic and immediate post-transition period, disruptions in health-care utilization, delayed diagnosis and treatment, catch-up of deferred care, and changes in reporting or registration may have affected burden estimates. The short-term fluctuation, characterized by increases from 2020 to 2022 followed by a decline in 2023, is consistent with special period effects during this interval. In line with this concern, the Joinpoint sensitivity analysis excluding data from 2020 to 2023 yielded lower AAPC estimates for incidence and mortality, indicating that trend estimates were influenced by this recent period. Continued monitoring of post-pandemic trends is therefore needed to distinguish transient period effects from sustained changes in bladder cancer burden.
The age-effect analysis showed that both incidence and mortality increased markedly with advancing age. Before 55 years of age, bladder cancer incidence and mortality remained relatively low and stable, possibly reflecting a shorter cumulative duration of carcinogenic exposure. After 55 years of age, both incidence and mortality increased substantially, reaching peaks at 85–89 years and 90–94 years, respectively. These findings further support bladder cancer as a strongly age-related malignancy, potentially associated with prolonged carcinogenic exposure, age-related immune decline, and reduced DNA damage repair capacity.35,36 Males consistently had higher incidence, mortality, and DALY rates than females. Several mechanisms may underlie this marked sex disparity. First, smoking, one of the most established risk factors for bladder cancer, is substantially more prevalent among men than among women in China.37 Second, men are more likely to be employed in occupations involving exposure to chemical carcinogens, such as aromatic amines, which may further increase bladder cancer risk.36 In addition, sex hormone signaling may contribute to the observed sex heterogeneity in bladder cancer risk. Previous studies suggest that androgen receptor signaling is more likely to promote bladder carcinogenesis and progression, whereas the role of estrogen receptor signaling appears to be subtype dependent, with estrogen receptor alpha showing more protective features and estrogen receptor beta more often associated with tumor-promoting effects.38,39
The long-term declines in age-standardized incidence, mortality, and DALY rates suggest that substantial progress has been made in bladder cancer control in China. However, increasing crude rates driven by population aging, recent unfavorable shifts in age-standardized trends, and projections indicating stable incidence with only modest declines in mortality through 2030 collectively indicate that the burden of bladder cancer remains substantial. These findings emphasize the need for continued optimization of prevention and control strategies. The large attributable fractions for smoking and high FPG observed in this study indicate that primary prevention should remain a major focus of bladder cancer control. Smoking remained the leading modifiable contributor to bladder cancer mortality, accounting for 38.06% of deaths overall and 46.60% among men, while high FPG accounted for 10.37% of deaths. Early detection and timely treatment also warrant further strengthening. Prognosis varies markedly by stage at diagnosis: the five-year relative survival rate is approximately 73% for localized disease, but declines to 41% for regional disease and 9% for distant disease. Together with the recent upward shifts in ASMR, these findings indicate that early detection and timely management of bladder cancer remain important areas for improvement. Cystoscopy, the diagnostic gold standard, is invasive and costly, which can reduce patient compliance and contribute to delayed evaluation. Although urine cytology is a common non-invasive adjunct in clinical practice, its utility is limited by suboptimal sensitivity, particularly for low-grade tumors.40 U.S. Food and Drug Administration-approved urine biomarkers, such as bladder tumor antigen Stat and bladder tumor antigen TRAK, NMP22, and UroVysion fluorescence in situ hybridization (FISH), have shown higher sensitivity than cytology in selected settings; however, routine use is constrained by false-positive results, variable diagnostic performance, and cost considerations.31 These challenges highlight the need for effective, non-invasive, and clinically accessible tools for early detection of BC. Moreover, the marked age effect and persistent sex disparity observed in this study underscore the need for future research on risk-stratified early-detection strategies. Older adults, men, smokers, and individuals with metabolic risk factors may represent priority groups for bladder cancer risk assessment using non-invasive early-detection tools.
Several limitations should be acknowledged. First, although GBD estimates integrate multiple high-quality national data sources and apply standardized modelling frameworks to improve representativeness and comparability, potential bias related to data-source heterogeneity, data incompleteness, and uncertainty propagation from model-based estimates is inevitable. Second, methodological limitations of the statistical framework warrant caution. Because the age-period-cohort model used the estimable-function framework, the estimated period and cohort rate ratios should be interpreted as identifiable relative risk functions reflecting temporal variation rather than fully isolated independent effects. In addition, because of computational constraints associated with iteratively refitting non-linear models (Joinpoint, APC, and projection models) across posterior distributions and limitations in data availability, downstream trend analyses were conducted using point estimates. Therefore, although 95% uncertainty intervals were provided for descriptive burden estimates, incomplete propagation of input uncertainty into the forecasting and segmentation models may have underestimated the overall statistical variance. Third, although potential effects of diagnostic practices, expanding registry coverage, and improvements in health-care access on observed burden trends were discussed, quantitative adjustment for these historical shifts was not feasible because of data limitations; thus, trend inflections should be interpreted cautiously. Fourth, although GBD risk-outcome pairs were selected through a rigorous and standardized process, this study could not fully capture the complete etiologic spectrum of bladder cancer. Finally, the ecological design of this study limits causal inference. Future studies incorporating indicators of diagnostic practice, health-care access, and detailed exposure histories are needed to clarify the drivers of recent trend inflections and better support risk-stratified prevention and early-detection strategies.