Global disparities in liver cancer burden: an epidemiological and socioeconomic analysis using GBD 1990-2021 data.

Introduction
Liver cancer, a heterogeneous group of malignancies primarily comprising hepatocellular carcinoma and intrahepatic cholangiocarcinoma, is a major global public health concern and remains one of the leading causes of cancer-related mortality worldwide. Recent global estimates indicate that liver cancer accounts for approximately 900,000 new cases and more than 800,000 deaths annually, with substantial variation in burden across geographic regions and socioeconomic settings [1]. Despite advances in cancer prevention and management, the global incidence and mortality of liver cancer have remained high, and the disease continues to rank among the top causes of cancer-related deaths worldwide [2, 3]. Liver cancer is characterized by a disproportionately high mortality-to-incidence ratio, reflecting its aggressive nature and the frequent diagnosis at advanced stages [4].
The global burden of liver cancer is closely linked to chronic liver diseases, including hepatitis B and C infections [5], alcohol-related liver disease [6], and metabolic dysfunction–associated steatotic liver disease (MASLD), formerly referred to as non-alcoholic fatty liver disease (NAFLD) [7, 8]. Marked geographic disparities have been consistently observed, with a disproportionate burden in regions such as Asia and sub-Saharan Africa, where viral hepatitis remains highly prevalent and access to effective prevention and treatment may be limited [9, 10]. In contrast, several high-income countries have experienced declining liver cancer burden over recent decades, partly attributable to successful hepatitis B vaccination programs and improved management of chronic viral hepatitis [11]. These contrasting patterns underscore the strong influence of socioeconomic context on liver cancer burden.
The Global Burden of Diseases, Injuries, and Risk Factors Study (GBD), a comprehensive epidemiological assessment coordinated by the Institute for Health Metrics and Evaluation (IHME), has been instrumental in providing detailed, country-level data on the burden of liver cancer worldwide. We utilized disability-adjusted life years (DALYs) as a metric to quantify the total burden of disease, capturing both the years of life lost due to premature mortality and the years lived with disability. Previous studies based on GBD data have demonstrated substantial geographic variation and temporal changes in liver cancer burden, as well as the important role of major etiological factors such as viral hepatitis, alcohol use, and metabolic disorders [12, 13].
However, despite these advances, a comprehensive and comparative assessment that simultaneously integrates long-term temporal trends, socioeconomic disparities, underlying demographic and epidemiological drivers, and health system performance across liver cancer subtypes remains limited. Importantly, the quantitative contribution of socioeconomic development relative to major etiologic risk factors has not been clearly delineated across the five SDI strata. Specifically, it remains uncertain whether observed SDI gradients are primarily explained by differences in risk factor attribution patterns, by demographic shifts, or by broader health system and development related factors captured by SDI. In particular, the extent to which population growth, population aging, and epidemiological changes differentially contribute to liver cancer burden across sociodemographic contexts has not been fully elucidated. Moreover, few studies have systematically evaluated the gap between observed liver cancer burden and the minimum achievable burden given a country’s level of socioeconomic development.
Therefore, this study used GBD 1990 to 2021 estimates to provide an updated assessment of liver cancer burden, including hepatocellular carcinoma and intrahepatic cholangiocarcinoma, at the global, regional, and national levels. To address the above knowledge gaps, this analysis aimed to quantify the magnitude and direction of liver cancer DALY disparities across the five SDI quintiles and to characterize their long-term temporal patterns using Joinpoint regression and age-period-cohort analyses. In parallel, changes in liver cancer DALYs were decomposed into contributions from population growth, population aging, and epidemiological changes, allowing comparison of how these components differ across SDI strata and liver cancer subtypes. The study also evaluated cross-country benchmarking performance by estimating the minimum observed burden conditional on SDI and identifying countries with the largest gaps between observed DALYs and the frontier. Finally, attribution to major risk factors, including hepatitis B, hepatitis C, alcohol use, and metabolic dysfunction–associated steatohepatitis(MASH), was examined across SDI strata to clarify how etiologic profiles relate to observed socioeconomic gradients in liver cancer burden. We hypothesized that SDI gradients in liver cancer burden reflect a combined influence of demographic change and heterogeneity in dominant etiologic risks, and that substantial unrealized potential for burden reduction exists in some countries even at similar levels of socioeconomic development.

Methods

Data sources
This was a population-based, comparative epidemiological and socioeconomic analysis using estimates from the Global Burden of Disease 2021 study [14, 15]. Specifically, this work represents a secondary longitudinal analysis of publicly available, aggregated and model-based GBD estimates rather than primary data collection. The unit of analysis was the country or territory year stratum, further disaggregated where applicable by sex, age group, and SDI quintile, and all inferences should be interpreted at the population level. All data were obtained from the Global Health Data Exchange (GHDx), the official data repository of the GBD study, using publicly available GBD 2021 results released by IHME.
A primary focus of this analysis was the burden of liver cancer quantified through DALYs. DALYs were age-standardized according to the GBD 2021 standard population to ensure comparability across countries and over time. Additionally, the research investigated the attributable burden linked to various risk factors, including alcohol consumption, hepatitis B, hepatitis C, and MASH. To enhance the understanding of these dynamics, the data were meticulously disaggregated by gender and SDI groups, which encompass five distinct categories: low SDI, low-middle SDI, middle SDI, high-middle SDI, and high SDI, as defined by the IHME SDI classification framework. The Sociodemographic Index is a composite measure of development constructed from lag-distributed income per capita, average educational attainment in the population aged 15 years and older, and the total fertility rate under age 25 years, and it is scaled from 0 to 1 to facilitate cross-country comparison. We followed the SDI definition and construction methods reported in the GBD methodological documentation and IHME technical materials.

Inclusion and exclusion criteria
In accordance with the GBD 2021 framework, this study included cases of primary liver cancer as defined by the GBD cause hierarchy, encompassing hepatocellular carcinoma, cholangiocarcinoma, hepatoblastoma, and other specified and unspecified primary malignant neoplasms of the liver. These disease categories were identified based on the corresponding International Classification of Diseases (ICD) codes as specified in the GBD 2021 cause mapping. This standardized case definition ensured consistency and comparability of liver cancer burden estimates across countries, time periods, and disease subtypes. The analysis focused exclusively on liver cancer burden quantified by DALYs.
Cases were excluded if they were secondary (metastatic) liver tumors originating from extrahepatic primary malignancies. In addition, benign liver tumors, non-malignant hepatic lesions, and liver involvement due to non-neoplastic diseases were not included in the analysis.
Burden estimates related to risk factors other than alcohol use, hepatitis B virus, hepatitis C virus, and MASH were excluded, as these were the primary etiological categories assessed in this study. Furthermore, no individual-level clinical or pathological data were included, as the analysis was conducted exclusively at the population level using standardized GBD estimates.

Spatial analysis
Spatial patterns of liver cancer burden were visualized using color-coded world maps depicting age-standardized DALY rates per 100,000 population. Maps were constructed using ArcGIS software (Esri, Redlands, CA), with country-level shapefiles obtained from the Global Administrative Areas database. Countries were classified using quantile-based categorization, and a sequential color scale was applied to reflect increasing burden levels, with darker colors representing higher DALY rates.

Age-period-cohort modeling
Age-period-cohort analysis was conducted to examine age effects, period effects, cohort effects, and overall temporal trends in liver cancer burden. The analysis followed established age-period-cohort modeling approaches commonly applied in Global Burden of Disease related studies and cancer epidemiology. Net drift, defined as the overall annual percentage change across all age groups, was estimated to summarize long-term temporal trends. In addition, age effects were derived from adjusted longitudinal age-specific rates, while period and cohort effects were expressed as relative risks compared with reference period and cohort categories. Statistical significance of net drift and effect estimates was assessed using Wald chi-squared tests. All APC analyses were conducted using R software, following standard implementations described in the epidemiological literature. Results are presented as estimable functions of age, period, and cohort effects, consistent with commonly accepted APC reporting practice.

Joinpoint regression analysis
Temporal trends in age-standardized liver cancer DALYs were assessed using Joinpoint regression analysis. A maximum of four joinpoints was allowed, and trend changes were evaluated using the Monte Carlo permutation test. Annual percent change and average annual percent change were estimated, and 95% confidence intervals were calculated for all AAPC estimates.

Decomposition analysis
Decomposition analysis was applied to quantify the relative contributions of population growth, population aging, and epidemiological changes to temporal variations in liver cancer DALYs. The analysis was conducted using the Das Gupta decomposition framework, which allows for additive partitioning of changes in disease burden into demographic and epidemiological components. Briefly, the change in total DALYs between two time points was expressed as the sum of three additive components attributable to population size, age structure, and age-specific DALY rates, respectively, following the standard Das Gupta formulation. This decomposition was performed consistently across SDI strata and liver cancer subtypes to facilitate direct comparison of driver contributions.

Frontier analysis
Frontier analysis was conducted to estimate the minimum achievable liver cancer DALYs conditional on a country’s level of socioeconomic development. A non-parametric locally estimated scatterplot smoothing (LOESS) approach was used to model the lower envelope of the relationship between SDI and age-standardized DALY rates. The resulting frontier represents the theoretical minimum burden observed across countries at comparable SDI levels, allowing identification of countries with substantial potential for improvement. Operationally, the frontier was defined as the estimated lower bound of observed DALY rates at a given SDI level, derived from the lower envelope of the SDI burden scatter using LOESS smoothing rather than the conditional mean. Countries were benchmarked by the absolute and relative distance between observed DALY rates and the frontier value at the same SDI level, and larger gaps were interpreted as greater unrealized potential for burden reduction within comparable development levels. Key LOESS settings and model specifications were recorded to support reproducibility, and sensitivity checks were performed using alternative smoothing parameters to confirm that the main benchmarking patterns were robust.

Results

Global burden and socioeconomic disparities in liver cancer
Geographic heterogeneity was observed in liver cancer burden in 2021, measured by DALYs. Countries such as Mongolia, Gambia, Mali, Eswatini, and Tonga had higher liver cancer DALY rates, whereas Mauritius, Kuwait, Argentina, Bermuda, and Morocco showed lower rates (Fig. 1A).

Temporal trend analyses demonstrated clear socioeconomic differences in overall liver cancer DALYs across SDI strata from 1990 to 2021 (Fig. 1B and G). At the global level, liver cancer DALYs declined over the study period, with an average annual percent change (AAPC) of − 0.46 (95% CI − 0.67 to − 0.26). Declining trends were observed in high SDI (AAPC − 0.42, 95% CI − 0.73 to − 0.12), high-middle SDI (AAPC − 0.63, 95% CI − 0.88 to − 0.37), middle SDI (AAPC − 0.72, 95% CI − 0.97 to − 0.47), and low SDI regions (AAPC − 0.80, 95% CI − 0.87 to − 0.74). In contrast, low-middle SDI countries experienced a modest increase in liver cancer DALYs, with an AAPC of 0.08 (95% CI 0.02 to 0.14). Country-specific joinpoint estimates are provided in Supplementary Data 1 and 2.
Decomposition analysis indicated that population growth and population aging were major contributors to changes in liver cancer DALYs globally between 1990 and 2021, while epidemiological changes contributed to a lesser extent (Fig. 1H).
Frontier analysis showed cross-country variability in the gap between observed liver cancer DALYs and the minimum achievable burden at comparable levels of socioeconomic development (Fig. 1I). In 2021, heterogeneity was observed in the frontier gap across countries. Among high SDI countries, Japan, Andorra, Monaco, Taiwan (Province of China), and the Republic of Korea showed larger gaps, whereas Somalia, Ethiopia, Haiti, Yemen, and Bangladesh showed smaller gaps (Fig. 1J).

Age-period-cohort patterns in liver cancer burden
Age–period–cohort analysis indicated heterogeneous temporal and generational patterns across SDI levels and sexes. The global net drift showed a modest increasing trend, whereas low and low-middle SDI countries showed stable or slightly declining net drifts. Middle and high-middle SDI regions showed higher net drift values, particularly among males (Fig. 2A).

Cohort effects were relatively flat at the global level but became more pronounced with increasing SDI, with high SDI regions showing a distinct cohort peak (Fig. 2B).
Period effects showed an overall decline globally, with faster reductions in high SDI regions than in low-middle SDI settings (Fig. 3A). Age effects followed an inverted U-shaped pattern, with peak burden occurring in middle-aged groups, and a more pronounced peak in low and low-middle SDI settings (Fig. 3B).

Summary of risk factor-specific patterns
Across major etiological categories, including alcohol use, hepatitis B, hepatitis C, MASH, and hepatoblastoma, several consistent patterns were observed. Liver cancer DALYs remained concentrated in low and low-middle SDI regions, temporal trends varied systematically across SDI levels, and demographic factors, especially population growth and aging, were major drivers of change. Frontier analyses consistently revealed variability in performance across countries at similar socioeconomic levels. Risk factor–specific results are presented below, with emphasis on distinct patterns.

Alcohol-attributable liver cancer
The geographic distribution of alcohol-attributable liver cancer DALYs across countries in 2021 showed higher DALY rates in Mongolia, Gambia, Mali, Eswatini, and Mozambique, and lower DALY rates in Mauritius, Morocco, Kuwait, Jordan, and Yemen (Fig. 4A).

From 1990 to 2021, alcohol-attributable liver cancer DALYs increased in low-middle SDI countries, with an AAPC of 0.63 (95% CI 0.43 to 0.82), and increased in high SDI countries, with an AAPC of 0.45 (95% CI 0.23 to 0.67). Decreases were observed in low SDI countries, with an AAPC of − 0.56 (95% CI − 0.63 to − 0.50). Trends in high-middle SDI countries were not statistically significant, with an AAPC of − 0.31 (95% CI − 0.75 to 0.13) (Fig. 4B and G). The detailed information is provided in Supplementary Data 3, 4.
Decomposition analysis indicated that population growth was a major contributor to changes in alcohol-attributable liver cancer DALYs in low and low-middle SDI settings, whereas population aging contributed more prominently in middle and higher SDI settings (Fig. 4H). Frontier analysis showed cross-country variability in frontier gaps in 2021 (Fig. 4I and J).

Hepatitis B-attributable liver cancer
The geographic distribution of hepatitis B–attributable liver cancer DALYs in 2021 showed higher DALY rates in Mongolia, Gambia, Mali, Guinea-Bissau, and Tonga and lower DALY rates in Morocco, Sweden, Argentina, Mexico, and El Salvador (Fig. 5A).

From 1990 to 2021, hepatitis B–attributable liver cancer DALYs declined in all SDI categories. The global AAPC was − 0.79 (95% CI − 0.99 to − 0.59). The largest decline was observed in high SDI countries, with an AAPC of − 1.18 (95% CI − 1.54 to − 0.82) (Fig. 5B and G). The detailed information is provided in Supplementary Data 5, 6.
Decomposition analysis showed that population growth was a major contributor in low to high-middle SDI settings, whereas epidemiological change was a primary contributor in high SDI settings (Fig. 5H). Frontier analysis showed cross-country variability in frontier gaps in 2021 (Fig. 5I and J).

Hepatitis C-attributable liver cancer burden
The geographic distribution of hepatitis C–attributable liver cancer DALYs in 2021 showed higher DALY rates in Mongolia, Egypt, Mali, Mozambique, and Zimbabwe and lower DALY rates in Mauritius, Morocco, Peru, Croatia, and Sri Lanka (Fig. 6A).

From 1990 to 2021, the global AAPC was − 0.31 (95% CI − 0.57 to − 0.05). Low-middle SDI countries showed a near-null trend that was not statistically significant, with an AAPC of 0.15 (95% CI − 0.0009 to 0.30). Declines were observed in low SDI countries, with an AAPC of − 0.59 (95% CI − 0.66 to − 0.52), and in high SDI countries, with an AAPC of − 0.49 (95% CI − 0.79 to − 0.20) (Fig. 6B and G). The detailed information is provided in Supplementary Data 7, 8.
Decomposition analysis indicated that population growth was a major contributor in low and low-middle SDI settings, while population aging contributed more in high-middle and high SDI settings (Fig. 6H). Frontier analysis showed cross-country variability in frontier gaps in 2021 (Fig. 6I and J).

MASH-attributable liver cancer burden
The geographic distribution of MASH-attributable liver cancer DALYs in 2021 showed higher DALY rates in Mongolia, Gambia, Mozambique, Mauritania, and Eswatini and lower DALY rates in Morocco, Mauritius, Argentina, Ukraine, and Haiti (Fig. 7A).

From 1990 to 2021, the global AAPC was 0.60 (95% CI 0.43 to 0.77). The largest increase was observed in low-middle SDI countries, with an AAPC of 0.99 (95% CI 0.89 to 1.10). In contrast, low SDI countries showed a decline, with an AAPC of − 0.16 (95% CI − 0.24 to − 0.08). Trends in high-middle SDI countries were not statistically significant, with an AAPC of 0.30 (95% CI − 0.04 to 0.65) (Fig. 7B and G). The detailed information is provided in Supplementary Data 9, 10.
Decomposition analysis indicated that population growth was a major contributor in low to middle SDI settings, while epidemiological changes contributed more in high SDI settings (Fig. 7H). Frontier analysis showed cross-country variability in frontier gaps in 2021 (Fig. 7I and J).

Hepatoblastoma and pediatric liver cancer burden
Hepatoblastoma represents a pediatric liver cancer subtype and showed distinct patterns compared with adult etiological categories. In 2021, higher DALY rates were observed in Guinea, Mongolia, Gambia, Mali, and Burkina Faso, while lower DALY rates were observed in Mauritius, Czechia, Croatia, Uruguay, and Sri Lanka (Fig. 8A).

From 1990 to 2021, hepatoblastoma DALYs declined across all SDI categories, with a global AAPC of − 2.37 (95% CI − 2.50 to − 2.23). The largest decline was observed in high-middle SDI countries, with an AAPC of − 4.13 (95% CI − 4.52 to − 3.74) (Fig. 8B and G). The detailed information is provided in Supplementary Data 11, 12.
Decomposition analysis indicated that epidemiological changes were the main contributors to the global decline, while demographic contributions varied by SDI category (Fig. 8H). Frontier analysis showed cross-country variability in frontier gaps in 2021 (Fig. 8I and J).

Discussion
The comprehensive global analysis of the liver cancer burden, as measured by DALYs, provides crucial insights into the complex and evolving patterns of this devastating disease across different socioeconomic and geographic contexts.
The geographic distribution of liver cancer DALYs demonstrates substantial global disparities. A markedly higher disease burden was observed in countries such as Mongolia, Gambia, Mali, Eswatini, and Tonga, whereas comparatively lower DALY rates were reported in Mauritius, Kuwait, Argentina, and Morocco. These differences indicate pronounced regional heterogeneity in liver cancer burden, potentially reflecting variations in underlying risk factor exposure, healthcare access, and prevention strategies. These regional differences may be largely explained by variations in the prevalence of major etiological factors, healthcare system capacity, and the implementation of effective prevention strategies. High-burden regions are often characterized by a high prevalence of chronic hepatitis B and C infections, limited access to antiviral therapy, suboptimal vaccination coverage, and delayed diagnosis, all of which substantially increase liver cancer risk [16, 17]. In contrast, countries with lower liver cancer DALYs generally benefit from earlier implementation of universal hepatitis B vaccination, stronger health systems, improved surveillance, and broader access to curative and palliative care services [18]. Importantly, our cross-national comparisons should be interpreted as reflecting differences in both underlying risk and health-system performance, while acknowledging that uncertainty may be higher in data-sparse locations. These results highlight where the burden is concentrated and provide an empirical basis for prioritizing settings where scaling proven interventions could yield the greatest absolute DALY reductions.
The temporal trends of liver cancer DALYs, stratified by SDI levels, reveal important insights. The steady increase in the low-middle SDI group suggests that these countries continue to face significant challenges in addressing the growing liver cancer burden. Conversely, the declining trend observed in the high SDI group is consistent with improved prevention and control capacity in these settings, including hepatitis B vaccination, earlier diagnosis, and expanded treatment availability [19, 20]. However, low SDI and middle SDI countries appear to be the most affected, requiring concerted efforts to mitigate the rising burden through enhanced public health initiatives, improved access to healthcare, and targeted interventions tailored to their specific needs [21, 22]. A key implication of the SDI-stratified trends is that the “same” intervention package may not produce comparable gains across contexts; baseline risk profiles and health-system capacity likely modify the achievable impact.
The decomposition analysis provides valuable insights into the underlying drivers of the observed changes in liver cancer DALYs. The predominant role of population growth and aging highlights the need for comprehensive strategies that address demographic shifts, while also tackling the epidemiological factors contributing to the disease burden [23, 24]. Therefore, interpreting temporal change requires distinguishing between compositional demographic shifts and changes in epidemiologic risk or care effectiveness. From a policy perspective, the decomposition results suggest that prevention and treatment improvements must outpace demographic expansion to achieve meaningful reductions in total DALYs, particularly in rapidly growing and aging populations.
The frontier analysis complements these findings by benchmarking the lowest attainable liver cancer DALYs at a given level of socioeconomic development. This approach suggests substantial potential for additional burden reduction, particularly in countries with outcomes far above the frontier. We interpret “distance to frontier” as a performance-benchmarking metric rather than a direct measure of efficiency or causal policy impact, but it can nonetheless help identify locations where outcomes appear worse than expected for their development level. In this way, frontier analyses can support priority setting by highlighting where improvement may be feasible even without immediate changes in SDI.
The specific countries identified as having the largest gaps between their observed liver cancer DALYs and the frontier present opportunities for impactful interventions. For instance, among the high SDI countries, Japan, Andorra, Monaco, Taiwan (Province of China), and the Republic of Korea have been highlighted as having significant potential for improvement. Conversely, Somalia, Ethiopia, Haiti, Yemen, and Bangladesh, among the low SDI countries, have been identified as making more progress in addressing their liver cancer burden. These insights can guide policymakers and public health authorities in these countries to develop tailored strategies, learn from best practices, and allocate resources effectively to further reduce the liver cancer burden.
The comprehensive analysis of the liver cancer burden attributable to various risk factors, such as alcohol use, hepatitis B, hepatitis C, and MASH, provides a nuanced understanding of the underlying drivers and their evolving dynamics across different socioeconomic contexts. These findings can inform the development of targeted interventions and public health policies that address the specific risk factors predominant in each region.
For instance, in countries with a high burden of liver cancer due to hepatitis B, such as Mongolia [25–27], Gambia [28, 29], and Mali [30], prioritizing high coverage hepatitis B immunization, improving linkage to antiviral treatment, and strengthening surveillance for chronic infection represent plausible, evidence informed strategies. The cited studies primarily contextualize local etiologic patterns; direct evidence on the magnitude of DALY reduction from specific intervention scale up in each country is more limited and may differ by baseline coverage, program quality, and health system capacity. Accordingly, these recommendations should be interpreted as policy relevant hypotheses that warrant confirmation using country specific implementation and evaluation data. Similarly, in settings with substantial hepatitis C attributable burden, such as Egypt and Mongolia, expanding screening and access to curative therapy is supported by a broader body of evidence, but the expected population level impact depends on feasible coverage, reinfection risk, and linkage to care in the current local context [31, 32].
In the case of liver cancer associated with MASH, the rising DALY in global underscores the need for comprehensive strategies that address the underlying drivers, such as the growing burden of obesity, diabetes, and metabolic disorders [33, 34]. Comprehensive strategies that integrate public awareness campaigns, the promotion of healthy lifestyle choices, and the incorporation of MASH screening and management into primary healthcare systems could play a significant role in addressing this emerging health threat. Such approaches offer a multifaceted response to liver disease prevention and could enhance early detection and intervention, ultimately reducing the burden of liver cancer [35, 36]. However, because attribution estimates are model-based and may vary in uncertainty across locations, the observed increases should be interpreted in conjunction with uncertainty intervals and local epidemiologic evidence where available.
Beyond population-level burden metrics, health-related quality of life (HRQOL) has become an increasingly important outcome in HCC, particularly in advanced disease. Recent systematic reviews indicate that systemic therapies, TACE, and surgical interventions can substantially influence patient-reported HRQOL, with therapeutic advances often preserving or improving quality of life even when survival benefits are modest. Studies have shown that various factors, including demographic, clinical, psychological, social, and symptomatic variables, can significantly affect HRQOL outcomes in patients with HCC.
For example, a systematic review on HRQOL in patients post-TACE highlights that symptom severity and psychological distress are consistently associated with lower HRQOL [37]. Another study focuses on the effects of systemic therapies, such as atezolizumab/bevacizumab and sorafenib, on HRQOL in advanced HCC patients. It concluded that although these therapies can worsen HRQOL compared to baseline, they show better HRQOL outcomes than older therapies like sorafenib [38]. Furthermore, a scoping review explored post-surgical HRQOL outcomes in HCC patients, revealing that liver transplantation generally results in substantial long-term HRQOL improvement, although challenges remain with immunosuppressive therapy [39].
The analysis of hepatoblastoma, a rare form of liver cancer primarily affecting children, highlights the importance of continued research and targeted interventions in this area [40, 41]. The observed downward trends, particularly in the high-middle SDI and high SDI groups, suggest that advancements in early diagnosis, specialized pediatric oncology care, and innovative treatment modalities may have contributed to the improved outcomes [42, 43]. Sustaining these efforts and ensuring equitable access to high-quality care for children with hepatoblastoma should remain a priority.

Strengths and limitations
This study has several strengths. It uses a standardized comparative framework to examine liver cancer DALYs across countries and over time, enabling coherent cross-setting comparisons. In addition, combining SDI-stratified trend analyses with decomposition and frontier approaches provides complementary perspectives on (i) how the burden changes, (ii) what components contribute to change, and (iii) how observed outcomes compare with an empirically derived benchmark at similar levels of development.
Several limitations should also be acknowledged. First, the analysis relies on GBD modelling assumptions, including the selection and harmonization of input data sources, covariate-informed estimation, and smoothing/borrowing strength across space and time. These assumptions improve comparability but may attenuate true local fluctuations, and in data-sparse settings the estimates may be more influenced by priors and covariate relationships, increasing uncertainty in levels and trends.
Second, underreporting and misclassification are likely more common in low-SDI countries due to limited diagnostic capacity, incomplete vital registration, and weaker routine health information systems. Such under-ascertainment could lead to underestimation of DALYs and may partially contribute to apparently lower rates in some settings; although GBD methods attempt to adjust for missingness, residual bias may remain.
Third, decomposition results are sensitive to model specification and reflect accounting of change under a chosen framework rather than causal attribution. Similarly, frontier models depend on functional-form assumptions and should be interpreted as benchmarking tools; “distance to frontier” does not directly quantify efficiency or the causal effect of any single policy. Finally, both decomposition and frontier analyses inherit measurement uncertainty from the underlying DALY estimates; therefore, interpretations should be considered alongside uncertainty intervals and the broader evidence base.
Fourth, the study is based on country level aggregated estimates, and thus is subject to ecological fallacy. Associations observed at the national level may not hold for individuals or for specific subpopulations, and they should not be interpreted as individual level causal effects.
Fifth, SDI is an aggregate index and can mask substantial within country heterogeneity. Subnational disparities in risk factor exposure, health system access, and case ascertainment may be large even in countries with similar national SDI values, which may dilute or obscure inequities relevant for policy.
Finally, where age period cohort modelling is used, APC estimates are constrained by the inherent identifiability problem, meaning that age, period, and cohort effects cannot be definitively separated. Accordingly, APC results should be interpreted as descriptive patterns under modelling assumptions rather than as unique causal components.
Generalizability of these findings to subnational settings and demographic subgroups is limited. National averages can conceal pronounced heterogeneity by region, urban rural residence, sex, socioeconomic position, and access to prevention and treatment services. Therefore, translating these results into policy should be complemented by subnational surveillance and locally disaggregated analyses where available, especially in settings with known geographic or social gradients in hepatitis prevalence, alcohol use, and metabolic risk.

Conclusion
This study demonstrates marked global disparities in the burden of liver cancer, with substantial heterogeneity across countries and sociodemographic development levels. High SDI regions generally experienced declining liver cancer DALY rates, whereas low and low-middle SDI countries showed increasing or persistently high burdens. Changes in liver cancer DALYs were driven by multiple factors, including demographic shifts such as population growth and aging, as well as differences in underlying risk profiles and health system capacity. These findings underscore that reductions in age-standardized rates alone may be insufficient to offset demographic pressures in many settings. The results highlight the importance of region-tailored prevention and control strategies that reflect local epidemiology and socioeconomic conditions, rather than uniform global approaches. Focusing on settings with the greatest disparities and understanding the dominant drivers of burden in each context will be critical for achieving meaningful and equitable reductions in liver cancer burden worldwide.

Supplementary Information