Epidemiological trends of NASH-related liver cancer in young and middle-aged populations: A global disease burden analysis from 1990 to 2021

Abstract

Recent decades have seen an increase in the incidence and mortality rates of early-onset cancers in adolescents and adults aged 15 to 49 years. The incidence and mortality rates of liver cancer caused by NAFLD have been increasing on an annual basis; however, the most recent global data on liver cancer caused by NAFLD has not been reported. A detailed analysis of trends in early-onset hepatocellular carcinoma attributable to NASH (EO-NRLC) from 1990 to 2021 was conducted using data from the global burden of disease study for 204 countries, 21 regions, and 5 sociodemographic indices (SDI) regions, with separate analysis conducted for each category. Prevalence rates, mortality rates, YLLs, YLDs, and disability-adjusted life years (DALYs) were calculated based on vital registration data, adjusted for age and reporting per 1,00,000 population, with a UI of 95%. Subsequently, we performed SDI analysis, as well as trend analysis based on age groups. In addition to this, decomposition analysis, age-period-cohort analysis, and Bayesian age-period-cohort projection and frontier analysis were also conducted. The prevalence (0.61 [0.49, 0.73]), mortality (0.48 [0.39, 0.58]), rate of DALYs (11.50 [9.39, 13.84]), and morbidity (0.49 [0.40, 0.60]) of EO-NRLCs in 2021 compared with 1990 were increasing. The analysis indicates that age-standardized indicators, encompassing prevalence, morbidity, mortality, and DALYs, have exhibited an accelerated increase in regions characterized by higher SDIs and more economically developed economies, particularly in Australia, Europe, and North America. Conversely, numerous low SDI countries and less developed economies continue to exhibit comparatively elevated age-standardized rates for these 4 indicators. It is also noteworthy that all indicators are higher for men than for women as a whole. Subsequent Bayesian age-period-cohort projection analyses suggest that the indicators will remain high and increasing through 2040. The substantial increase in the incidence, prevalence, mortality, and DALYs of EO-NRLC across multiple regions from 1990 to 2021 is projected to impose a considerable socioeconomic burden on governments and health systems in the coming years. The findings of this study may inform policymakers in developing strategies to address EO-NRLC, with particular emphasis on strengthening professional training to mitigate the burden of this complex disease.

1. Introduction

Conventionally, cancer has been viewed as a disease primarily affecting older adults, typically those aged 50 and above. However, emerging evidence highlights a significant rise in early-onset cancers – defined as malignancies diagnosed in individuals under 50 – across various organ systems.^[1,2] This trend underscores the need to recognize the unique healthcare challenges faced by this younger population, including barriers to screening, academic and occupational pressures, adverse lifestyle factors, and socioeconomic constraints.^[3,4] Addressing these challenges requires timely adoption of precise diagnostics, personalized treatment strategies, and comprehensive care approaches to mitigate the growing burden of early-onset cancer.^[5]

To alleviate the cancer burden, urgent implementation of timely diagnosis, tailored treatments, and comprehensive care is required. The natural course of NAFLD progresses from simple steatosis (NAFL) to steatohepatitis (nonalcoholic steatohepatitis [NASH]), and can lead to advanced complications like cirrhosis and hepatocellular carcinoma (HCC).^[6,7] While NAFLD-associated HCC remains less common than HCV-related HCC, NAFLD is the most prevalent liver disease worldwide. Its associated HCC incidence is rising significantly, paralleling the global obesity epidemic.^[8,9] Emerging evidence suggests NAFLD may be a distinct, potentially modifiable risk factor for gastrointestinal cancers in younger individuals.^[10] Therefore, investigating trends in early-onset HCC is critically important.

The Global Burden of Disease (GBD) database serves as a key resource for quantifying the health impacts of diseases, injuries, and risk factors worldwide.^[11] By analyzing GBD data, this study investigates regional and temporal disparities in the incidence of early-onset HCC due to NASH (EO-NRLC). The findings provide insights into the optimal allocation of healthcare resources and support the development of updated, region-specific prevention and management strategies.

2. Materials and methods

2.1. Data source

The epidemiological data pertaining to HCC associated with NASH were extracted from the GBD 2021 repository. This investigation focused on analyzing the disease burden of NASH-related hepatic malignancies (NASH) within the 15 to 49 age demographic, utilizing comprehensive datasets from the GBD 2021 study.^[12] This comprehensive repository provides systematic evaluations of adverse health outcomes associated with 371 distinct pathological conditions, traumatic injuries, and functional impairments, in addition to 88 unique risk determinants. Spanning 204 geographical entities, the database incorporates contemporary epidemiological evidence and sophisticated standardized analytical frameworks to generate precise and up-to-date assessments of global health patterns.^[13] The present investigation employed a systematic analytical framework to quantify multiple health metrics, encompassing disease prevalence, incidence rates, case fatality, and DALY measurements. This methodological framework guaranteed standardized measurement and longitudinal comparability across diverse geographical regions and temporal dimensions. The study further examined the incidence of prevalence, mortality, disability-adjusted life years (DALYs), and age-standardized rates (ASRs) across 21 districts, including 5 districts with varying levels of SocioDemographic Index (SDI). Subsequently, we visualized and contrasted the regions in 1990 and 2021 by utilizing the ASRs for each of the 4 indicators across 204 countries and territories. This study employed publicly available data from the GBD study and, as such, did not necessitate ethical approval. All methodologies were executed in full compliance with the applicable standards and regulatory frameworks.

2.2. Statistical analysis

In this study, we employed connected point regression to detect significant temporal shifts in trends. A joinpoint model was constructed to estimate the inflection points where trends alter in direction or magnitude, utilizing Joinpoint software (version 5.2.0; National Cancer Institute). The determination of statistical significance for individual joinpoints was conducted with a 95% confidence level, enabling computation of both annual percentage variation and mean annual percentage variation metrics. The model initially fits a linear trend, subsequently aligns the data to identify potential inflection points, and ultimately computes the AAPC as a weighted mean of the APC for each segment. All statistical analyses and data visualizations were performed using R (version 4.4.2) and JD_GBDR (V2.37, Jingding Medical Technology Co., Ltd.). In this study, the R software package (version 4.2.3) and JD_GBDR (V2.22, Jingding Medical Technology Co., Ltd.) was used for the drawing of the figures.

Statistical significance was established at P < .05 for all inferential analyses.

2.3. Evaluation of socioeconomic development indicators

The socioeconomic development index represents an essential metric for evaluating regional variations in developmental status and economic well-being. This composite metric integrates elements such as gross national income per capita, mean years of educational attainment, and national fertility rates, providing a comprehensive overview of socioeconomic development at the national scale. Elevated SDI values denote a higher degree of socioeconomic progress. Based on the calculated SDI scores, global regions are systematically classified into 5 distinct strata: low, lower-middle, middle, upper-middle, and high SDI categories, reflecting varying levels of development. In this research, we examined the varying SDI levels of 4 specific indicators across 21 regions and performed a correlation analysis to explore their relationships.

2.4. Gender regional differences and population analysis

Epidemiological data were extracted from the GBD repository to analyze demographic variations stratified by gender and geographical distribution within 5 socioeconomic development quintiles, with particular emphasis on 4 key parameters. Furthermore, we conducted a correlation analysis across various gender and age cohorts to uncover potential interrelationships.

2.5. Risk factors

he GBD 2021 investigation delineated 5 critical etiological determinants associated with the development of EO-NRLC. These include tobacco smoking, hyperglycemia, excessive alcohol consumption, inadequate physical activity, tobacco use, and various metabolic dysfunctions. The analytical approaches and conceptual paradigms utilized for evaluating the epidemiological influence of these etiological determinants on EO-NRLC burden have been thoroughly investigated in existing scientific literature. Utilizing a validated risk quantification framework, we determined the proportional mortality and DALY metrics associated with individual etiological factors across different age strata for EO-NRLC.

2.6. Decomposition analysis

Discretization analysis is an indispensable tool for uncovering the complex factors behind the changing incidence, mortality and other health indicators of diseases over time. This approach dissects the overall variations into 3 fundamental elements: epidemiological transition, population expansion, and demographic aging. This analytical approach measures the proportional impact of individual determinants, providing comprehensive insights into temporal variations in non-communicable disease burden across the observation period. This methodological framework enables the identification of optimal public health priorities, determining whether resource allocation should emphasize risk factor reduction, population demographic management, or integrated approaches, thus informing the development of precision-based intervention strategies.

2.7. Age-period-cohort analysis

To investigate the impact of age, temporal period, and birth cohort on the burden of early-onset non-rheumatic left ventricular cardiomyopathy (EO-NRLC), an age-period-cohort (APC) analytical approach was implemented. This analytical approach enables the dissection of intricate interactions among these factors by isolating their individual contributions, thus providing a clearer understanding of their distinct effects. This methodology utilizes the APC model, which effectively controls for confounding variables to provide a more precise evaluation of trends. In particular, the age effect examines the differential impact of period-specific determinants on the burden of EO-NRLC. Meanwhile, cohort effects underscore variations in disease patterns across distinct birth cohorts, elucidating the influence of intergenerational factors. This approach provides a comprehensive and nuanced framework for understanding how demographic aging, temporal fluctuations, and intergenerational dynamics collectively shape the progression of disease burden across time.

2.8. Predictive analytics

To forecast the disease burden of EO-NRLC from 2022 to 2040 in the absence of intervention, a Bayesian APC model incorporating nested Laplace approximations was utilized.^[14]

2.9. Frontier analysis

A frontier analysis methodology was employed to identify the optimal benchmarks for assessing the disease burden within the EO-NASH-LC. This was accomplished by evaluating the performance of individual countries and regions against those exhibiting the highest levels of achievement. For every country and region, the “effective margin” was determined by calculating the disparity between the actual disease burden of EO-NRLC and the potential burden that could be realized, based on their respective SDIs. This was undertaken to pinpoint the leading countries, positioning them as exemplars for others to emulate.

3. Results

3.1. Comparison of EO-NRLC burdens in global regions, districts

3.1.1. Global trends in burden of EO-NRLC, 1990–2021

Overall, the global burden of early-onset NASH-related liver cancer (EO-NRLC) increased steadily from 1990 to 2021. As shown in Tables 1–3,The number of prevalent cases more than doubled, rising from 3447 (95% UI: 2671–4345) in 1990 to 7233 (95% UI: 5608–9155) in 2021 (AAPC: 0.007; 95% CI: 0.007–0.007).

Table 1.

Number of deaths due to EO-NRLC and ASMR (per 1,00,000 population) 95% confidence intervals for 1990 and 2021 and AAPC (1990–2021).

Characteristics	1990		2021		1990–2021
	Number (95% UI)	Rate per 1,00,000 no. (95% UI)	Number (95% UI)	Rate per 1,00,000 no. (95% UI)	AAPCs% (95% CI) 1990–2021
Global	1989.62 (1537.69, 2559.48)	0.38 (0.30, 0.47)	3545.07 (2729.36, 4557.85)	0.48 (0.39, 0.58)	0.003 (0.003, 0.004)
Sex
Female	930.38 (720.21, 1183.26)	0.36 (0.28, 0.46)	1609.32 (1260.02, 2055.90)	0.44 (0.36, 0.54)	0.003 (0.003, 0.003)
Male	1059.24 (796.49, 1404.66)	0.39 (0.31, 0.49)	1935.76 (1442.36, 2627.73)	0.52 (0.41, 0.64)	0.004 (0.004, 0.004)
SDI
High SDI	206.96 (154.64, 268.19)	0.30 (0.24, 0.38)	276.32 (206.25, 358.53)	0.44 (0.35, 0.55)	0.004 (0.004, 0.004)
High-middle SDI	442.28 (337.35, 555.43)	0.32 (0.26, 0.38)	557.26 (423.26, 728.58)	0.38 (0.30, 0.46)	0.002 (0.002, 0.002)
Low SDI	211.08 (138.15, 324.94)	0.70 (0.44, 1.09)	511.82 (334.51, 771.30)	0.68 (0.48, 0.94)	−0.001 (−0.001, −0.001)
Low-middle SDI	331.52 (243.20, 459.15)	0.36 (0.26, 0.52)	893.36 (665.06, 1147.84)	0.50 (0.40, 0.62)	0.004 (0.004, 0.004)
Middle SDI	796.65 (624.68, 995.05)	0.46 (0.37, 0.56)	1304.39 (1009.86, 1686.43)	0.52 (0.42, 0.64)	0.002 (0.002, 0.003)
Reigon
Andean Latin America	6.34 (4.15, 9.16)	0.29 (0.20, 0.41)	12.73 (7.88, 19.54)	0.37 (0.24, 0.53)	0.002 (0.002, 0.003)
Australasia	3.02 (2.12, 4.15)	0.15 (0.11, 0.20)	13.19 (9.30, 18.36)	0.55 (0.39, 0.74)	0.013 (0.012, 0.013)
Caribbean	4.88 (3.23, 6.90)	0.21 (0.14, 0.29)	8.00 (5.04, 12.21)	0.22 (0.15, 0.31)	0.000 (0.000, 0.001)
Central Asia	27.22 (18.24, 39.23)	0.54 (0.37, 0.76)	46.47 (29.86, 67.72)	0.61 (0.41, 0.87)	0.002 (0.002, 0.003)
Central Europe	23.60 (15.97, 33.93)	0.28 (0.19, 0.38)	19.07 (13.25, 26.70)	0.29 (0.21, 0.39)	0.000 (0.000, 0.001)
Central Latin America	25.51 (19.23, 32.44)	0.26 (0.20, 0.34)	63.65 (49.12, 81.44)	0.33 (0.26, 0.42)	0.002 (0.001, 0.002)
Central Sub-Saharan Africa	20.32 (7.90, 48.84)	0.54 (0.21, 1.23)	46.09 (17.18, 108.27)	0.46 (0.18, 1.12)	−0.003 (−0.003, −0.003)
East Asia	850.24 (655.34, 1084.13)	0.50 (0.40, 0.61)	960.76 (710.91, 1267.45)	0.51 (0.40, 0.64)	0.001 (0.001, 0.002)
Eastern Europe	29.66 (24.14, 35.81)	0.14 (0.12, 0.16)	45.76 (35.70, 56.10)	0.23 (0.19, 0.27)	0.003 (0.002, 0.003)
Eastern Sub-Saharan Africa	80.22 (56.60, 113.31)	0.76 (0.52, 1.10)	212.57 (139.02, 318.85)	0.80 (0.55, 1.11)	0.001 (0.001, 0.001)
High-income Asia Pacific	94.72 (66.68, 133.72)	0.71 (0.56, 0.91)	42.09 (28.01, 62.79)	0.48 (0.37, 0.62)	−0.007 (−0.008, −0.006)
High-income North America	41.22 (33.91, 48.57)	0.21 (0.18, 0.24)	89.21 (72.16, 106.46)	0.49 (0.41, 0.58)	0.009 (0.009, 0.009)
North Africa and Middle East	95.98 (61.08, 148.59)	0.44 (0.27, 0.74)	323.09 (216.38, 453.07)	0.69 (0.47, 0.94)	0.008 (0.008, 0.009)
Oceania	1.89 (0.93, 3.73)	0.39 (0.21, 0.80)	4.00 (2.16, 7.57)	0.36 (0.20, 0.62)	−0.001 (−0.001, −0.001)
South Asia	210.89 (171.36, 257.68)	0.24 (0.19, 0.29)	626.97 (494.55, 773.96)	0.38 (0.32, 0.45)	0.004 (0.004, 0.005)
Southeast Asia	221.13 (160.21, 298.73)	0.60 (0.44, 0.82)	459.41 (301.13, 670.57)	0.69 (0.48, 0.94)	0.003 (0.002, 0.003)
Southern Latin America	1.90 (1.20, 2.83)	0.07 (0.05, 0.10)	6.90 (4.56, 10.00)	0.19 (0.13, 0.27)	0.004 (0.004, 0.004)
Southern Sub-Saharan Africa	40.85 (25.55, 61.30)	0.67 (0.40, 1.01)	106.76 (78.63, 140.58)	1.15 (0.92, 1.42)	0.015 (0.014, 0.016)
Tropical Latin America	17.40 (14.40, 20.54)	0.13 (0.11, 0.15)	33.39 (27.42, 39.60)	0.17 (0.14, 0.19)	0.001 (0.001, 0.001)
Western Europe	45.64 (33.57, 59.85)	0.20 (0.15, 0.27)	74.66 (54.22, 99.72)	0.32 (0.23, 0.42)	0.004 (0.004, 0.004)
Western Sub-Saharan Africa	147.00 (83.68, 248.74)	1.27 (0.72, 2.16)	350.28 (229.16, 517.72)	1.23 (0.91, 1.65)	−0.002 (−0.002, −0.001)

ASMR = age-standardized mortality, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

Table 3.

Number of deaths due to EO-NRLC and ASPR (per 1,00,000 population) 95% confidence intervals for 1990 and 2021 and AAPC(1990–2021).

Characteristics	1990	Rate per 1,00,000 no. (95% UI)	2021	Rate per 1,00,000 no. (95% UI)	1990–2021
	Number (95% UI)		Number (95% UI)		AAPCs% (95% CI) 1990–2021
Global	3446.64 (2670.90, 4345.12)	0.40 (0.32, 0.49)	7233.42 (5608.35, 9155.02)	0.61 (0.49, 0.73)	0.007 (0.007, 0.007)
Sex
Female	1630.92 (1269.06, 2070.36)	0.37 (0.30, 0.47)	3313.72 (2583.82, 4192.97)	0.59 (0.44, 0.77)	0.005 (0.005, 0.005)
Male	1815.72 (1394.38, 2349.61)	0.47 (0.36, 0.59)	3919.70 (2964.86, 5169.80)	0.78 (0.56, 1.05)	0.008 (0.008, 0.009)
SDI
High SDI	479.96 (372.31, 604.55)	0.41 (0.33, 0.50)	1022.02 (782.92, 1310.17)	0.82 (0.67, 1.03)	0.013 (0.013, 0.013)
High-middle SDI	753.31 (587.53, 938.05)	0.62 (0.40, 0.94)	1254.83 (955.51, 1618.27)	0.61 (0.43, 0.83)	0.005 (0.005, 0.005)
Low SDI	341.28 (220.16, 524.23)	0.44 (0.36, 0.53)	850.98 (556.00, 1287.08)	0.59 (0.46, 0.71)	−0.001 (−0.001, −0.000)
Low-middle SDI	543.72 (399.61, 745.54)	0.34 (0.25, 0.47)	1511.49 (1132.47, 1924.99)	0.48 (0.38, 0.59)	0.004 (0.004, 0.005)
Middle SDI	1326.48 (1065.03, 1643.55)	0.33 (0.27, 0.39)	2590.59 (2002.34, 3339.81)	0.47 (0.37, 0.57)	0.005 (0.005, 0.005)
Reigon
Andean Latin America	10.65 (6.98, 15.25)	0.25 (0.18, 0.35)	22.80 (14.18, 34.32)	0.32 (0.21, 0.46)	0.002 (0.002, 0.003)
Australasia	6.79 (4.84, 9.14)	0.18 (0.14, 0.23)	45.08 (31.75, 61.56)	0.90 (0.67, 1.21)	0.024 (0.023, 0.024)
Caribbean	8.34 (5.54, 11.73)	0.19 (0.14, 0.26)	14.24 (8.99, 21.26)	0.22 (0.15, 0.30)	0.001 (0.001, 0.001)
Central Asia	45.78 (30.81, 66.85)	0.51 (0.36, 0.71)	76.65 (49.99, 112.56)	0.56 (0.39, 0.80)	0.001 (0.001, 0.002)
Central Europe	38.52 (25.65, 55.00)	0.24 (0.17, 0.34)	32.34 (22.88, 45.16)	0.28 (0.20, 0.37)	0.001 (0.001, 0.001)
Central Latin America	43.03 (32.83, 54.66)	0.23 (0.18, 0.30)	113.06 (87.32, 145.61)	0.30 (0.24, 0.38)	0.002 (0.002, 0.002)
Central Sub-Saharan Africa	33.01 (12.91, 78.52)	0.49 (0.21, 1.07)	75.60 (28.78, 174.32)	0.41 (0.17, 0.95)	−0.003 (−0.003, −0.003)
East Asia	1425.19 (1094.36, 1800.69)	0.51 (0.41, 0.62)	2246.35 (1681.04, 2938.78)	0.68 (0.52, 0.86)	0.007 (0.006, 0.007)
Eastern Europe	49.42 (40.42, 59.74)	0.14 (0.12, 0.16)	78.18 (61.97, 95.22)	0.22 (0.19, 0.26)	0.003 (0.002, 0.003)
Eastern Sub-Saharan Africa	130.06 (90.91, 182.33)	0.67 (0.46, 0.95)	350.32 (228.37, 519.44)	0.71 (0.49, 0.97)	0.001 (0.001, 0.001)
High-income Asia Pacific	217.21 (160.75, 290.81)	1.09 (0.89, 1.33)	212.67 (142.33, 314.44)	1.23 (0.95, 1.58)	0.001 (−0.000, 0.003)
High-income North America	119.98 (98.42, 142.04)	0.27 (0.24, 0.31)	375.30 (307.06, 446.66)	0.86 (0.73, 1.00)	0.019 (0.018, 0.019)
North Africa and Middle East	159.44 (102.93, 246.16)	0.40 (0.25, 0.65)	580.26 (393.56, 816.29)	0.66 (0.47, 0.91)	0.009 (0.008, 0.009)
Oceania	3.07 (1.54, 5.98)	0.37 (0.20, 0.73)	6.67 (3.59, 12.05)	0.34 (0.19, 0.61)	−0.001 (−0.001, −0.001)
South Asia	342.98 (281.16, 419.25)	0.23 (0.19, 0.28)	1055.29 (835.74, 1300.57)	0.36 (0.30, 0.42)	0.004 (0.004, 0.004)
Southeast Asia	361.57 (264.36, 494.35)	0.57 (0.41, 0.78)	809.00 (530.73, 1175.55)	0.70 (0.49, 0.95)	0.004 (0.004, 0.004)
Southern Latin America	3.17 (2.01, 4.65)	0.06 (0.04, 0.09)	13.46 (8.96, 19.40)	0.19 (0.13, 0.27)	0.004 (0.004, 0.004)
Southern Sub-Saharan Africa	71.63 (45.10, 105.97)	0.63 (0.39, 0.95)	179.84 (132.45, 237.50)	1.04 (0.83, 1.28)	0.013 (0.012, 0.014)
Tropical Latin America	29.13 (24.26, 34.23)	0.12 (0.10, 0.14)	57.76 (47.89, 68.43)	0.16 (0.13, 0.18)	0.001 (0.001, 0.001)
Western Europe	106.41 (79.69, 136.63)	0.23 (0.17, 0.29)	294.44 (216.83, 388.60)	0.54 (0.41, 0.71)	0.010 (0.010, 0.010)
Western Sub-Saharan Africa	241.27 (137.49, 402.59)	1.13 (0.67, 1.88)	594.12 (390.74, 877.93)	1.08 (0.80, 1.44)	−0.002 (−0.003, −0.002)

ASPR = age-standardized prevalence, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

Table 2.

Number of deaths due to EO-NRLC and ASDR (per 1,00,000 population) 95% confidence intervals for 1990 and 2021 and AAPC (1990–2021).

Characteristics	1990	Rate per 1,00,000 no. (95% UI)	2021	Rate per 1,00,000 No. (95% UI)	1990–2021
	Number (95% UI)		Number (95% UI)		AAPCs% (95% CI) 1990–2021
Global	1,02,413.18 (79,250.36, 129,214.46)	9.63 (7.66, 11.90)	1,79,340.89 (1,40,049.36, 2,28,424.25)	11.50 (9.39, 13.84)	0.062 (0.059, 0.066)
Sex
Female	49,157.38 (38,240.21, 62,096.87)	9.01 (7.21, 11.36)	83,561.99 (65,774.07, 1,05,787.50)	10.48 (8.57, 12.63)	0.048 (0.045, 0.052)
Male	53,255.80 (40,611.47, 69,490.96)	10.25 (8.11, 12.66)	95,778.90 (71,863.43, 1,28,061.42)	12.59 (9.89, 15.72)	0.078 (0.074, 0.081)
SDI
High SDI	10,266.84 (7830.69, 13,289.14)	7.29 (5.88, 9.15)	13,724.03 (10,452.62, 17,640.19)	9.62 (7.83, 11.98)	0.074 (0.068, 0.079)
High-middle SDI	22,447.89 (17,432.04, 27,956.74)	8.36 (6.90, 10.02)	27,059.96 (20,579.88, 35,297.68)	9.11 (7.20, 10.93)	0.027 (0.021, 0.033)
Low SDI	11,150.23 (7267.21, 17,069.26)	16.88 (10.93, 25.58)	27,242.68 (17,728.56, 41,078.90)	16.13 (11.52, 22.31)	−0.031 (−0.034, −0.027)
Low-middle SDI	17,544.80 (12,975.58, 24,091.03)	9.19 (6.84, 12.67)	46,602.96 (35,363.31, 59,356.09)	12.55 (10.02, 15.42)	0.110 (0.107, 0.113)
Middle SDI	40,944.76 (32,839.81, 51,018.85)	11.69 (9.50, 14.14)	64,614.36 (50,178.51, 82,692.84)	12.49 (9.94, 15.18)	0.034 (0.028, 0.040)
Reigon
Andean Latin America	344.88 (226.58, 490.28)	6.51 (4.69, 8.92)	669.07 (413.36, 1014.28)	7.80 (5.09, 11.18)	0.041 (0.026, 0.055)
Australasia	156.48 (109.55, 212.09)	3.74 (2.82, 4.87)	657.90 (468.56, 903.78)	12.66 (9.28, 16.87)	0.287 (0.281, 0.292)
Caribbean	253.48 (168.65, 350.71)	4.76 (3.42, 6.51)	402.19 (255.61, 605.39)	5.15 (3.55, 6.97)	0.011 (0.006, 0.016)
Central Asia	1456.66 (993.13, 2078.36)	13.80 (9.72, 19.27)	2374.52 (1558.57, 3506.27)	14.71 (10.10, 20.96)	0.021 (0.012, 0.031)
Central Europe	1203.90 (812.45, 1705.11)	6.27 (4.53, 8.53)	928.66 (657.50, 1291.12)	6.43 (4.73, 8.69)	0.011 (0.005, 0.016)
Central Latin America	1375.31 (1045.83, 1737.31)	6.05 (4.73, 7.73)	3346.34 (2587.97, 4284.00)	7.47 (5.91, 9.20)	0.049 (0.039, 0.059)
Central Sub-Saharan Africa	1078.96 (431.58, 2557.93)	13.59 (5.81, 29.71)	2432.85 (940.60, 5668.16)	11.10 (4.55, 25.62)	−0.083 (−0.087, −0.079)
East Asia	43,024.89 (33,272.99, 54,267.95)	13.41 (10.84, 16.50)	46,367.98 (34,722.30, 61,017.90)	12.38 (9.61, 15.55)	−0.014 (−0.030, 0.001)
Eastern Europe	1560.21 (1284.20, 1863.43)	3.57 (3.03, 4.18)	2280.93 (1812.31, 2767.46)	5.45 (4.65, 6.34)	0.057 (0.048, 0.067)
Eastern Sub-Saharan Africa	4305.88 (2990.64, 5987.84)	18.42 (12.88, 25.72)	11,331.43 (7439.92, 16,900.71)	18.91 (12.89, 26.16)	0.010 (0.005, 0.014)
High-income Asia Pacific	4524.73 (3178.68, 6304.08)	17.50 (13.96, 21.91)	1977.64 (1324.54, 2941.88)	9.88 (7.67, 12.91)	−0.224 (−0.251, −0.198)
High-income North America	2119.62 (1749.46, 2485.51)	4.63 (3.98, 5.30)	4550.79 (3748.13, 5388.56)	10.98 (9.33, 12.85)	0.207 (0.202, 0.212)
North Africa and Middle East	5058.90 (3283.13, 7840.56)	10.58 (6.71, 17.47)	16,557.71 (11,286.78, 23,220.21)	16.43 (11.62, 22.51)	0.188 (0.178, 0.198)
Oceania	98.79 (49.27, 197.00)	9.89 (5.39, 19.76)	206.75 (111.64, 384.75)	8.83 (5.02, 16.23)	−0.037 (−0.039, −0.034)
South Asia	11,210.07 (9200.73, 13,626.02)	6.27 (5.12, 7.50)	32,753.62 (25,961.97, 40,509.37)	9.34 (7.93, 11.07)	0.102 (0.098, 0.106)
Southeast Asia	11,363.43 (8368.54, 15,297.22)	15.18 (10.97, 20.62)	22,644.22 (14,902.98, 32,872.47)	16.41 (11.55, 22.39)	0.042 (0.035, 0.049)
Southern Latin America	94.98 (60.18, 140.40)	1.61 (1.12, 2.24)	343.37 (232.26, 489.68)	4.32 (2.99, 5.92)	0.086 (0.082, 0.089)
Southern Sub-Saharan Africa	2193.95 (1381.12, 3282.30)	16.94 (10.36, 25.50)	5534.76 (4076.05, 7286.09)	27.81 (22.24, 34.31)	0.336 (0.308, 0.364)
Tropical Latin America	917.46 (769.21, 1072.28)	3.16 (2.76, 3.59)	1709.32 (1425.52, 2020.08)	3.96 (3.36, 4.56)	0.028 (0.024, 0.031)
Western Europe	2352.78 (1755.55, 3043.52)	4.57 (3.51, 5.87)	3718.21 (2740.91, 4894.29)	6.95 (5.25, 9.12)	0.075 (0.072, 0.078)
Western Sub-Saharan Africa	7717.81 (4420.32, 12,879.34)	30.56 (17.91, 50.58)	18,552.63 (12,192.28, 27,137.90)	28.33 (20.95, 37.63)	−0.089 (−0.103, −0.075)

ASDR = age-standardized death rate, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

Mortality also rose significantly. Deaths increased from 1990 (95% UI: 1537–2559) in 1990 to 3545 (95% UI: 2272–4558) in 2021, with the age-standardized mortality rate (ASMR) climbing from 0.38 to 0.48/1,00,000 (AAPC: 0.05, P < .05). Similarly, DALYs grew from 1,02,413 (95% UI: 79,250–1,29,214) to 1,79,341 (95% UI: 1,40,049–2,28,424) over the same period (AAPC: 0.062; 95% CI: 0.059–0.066).

3.2. Geographic and sociodemographic variations

A sociodemographic index (SDI)-based analysis revealed heterogeneous trends. Among the 5 SDI regions, only the low-SDI region showed a significant decrease in prevalence (AAPC: −0.001) and mortality (AAPC: −0.001), alongside a decline in DALYs (AAPC: −0.031). In contrast, all other SDI regions experienced increases in these burden metrics.

At the regional level, age-standardized DALYs decreased in the high-income Asia–Pacific, Central Sub-Saharan Africa, East Asia, Oceania, and Western Sub-Saharan Africa. Reductions in mortality were observed only in Western Sub-Saharan Africa, while prevalence declined in Central Sub-Saharan Africa, Oceania, and Western Sub-Saharan Africa. All other global regions saw an increase in burden indicators.

3.3. Gender disparity

The rising trend was consistent across genders, but men consistently bore a greater disease burden. The male proportion of total EO-NRLC cases increased from 52.67% in 1990 to 54.19% in 2021.

3.4. Temporal patterns of age-standardized rates

ASRs exhibited distinct temporal patterns (Fig. 1):

Figure 1.

Trends in global EO-NRLC incidence, prevalence, mortality, and DALY rates from 1990 to 2021. (A) Mortality, (B) disability-adjusted life years (DALYs), (C) prevalence, and (D) incidence. DALY = disability-adjusted life year, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

Mortality and DALY rates (ASMR and age-standardized death rates [ASDR]): Both h showed an initial rise before 2000, a decline between 2000 and 2005, and a resumed upward trend until a slight downturn after 2016.

Prevalence and incidence rates (age-standardized prevalence rates (ASPR) and age-standardized incidence rates [ASIR]): In contrast, the ASPR demonstrated a near-consistent increase across the study period, with only a minor slowdown after 2015. The ASIR followed an overall upward trajectory, with brief periods of slow decline (2000–2005 and post-2015), yet remained at elevated levels.

3.5. Geographic distribution shifts

Maps comparing 1990 and 2021 reveal notable shifts in the geographic distribution of EO-NRLC burden (Fig. 2).

Figure 2.

The geographic distribution of the age-standardized rate (ASR) for early-onset nonalcoholic fatty liver cancer (EO-NRLC) in 1990 and 2021 covered 204 countries and regions. (A) Trends in ASR for mortality from 1990 (left) to 2021 (right). (B) Trends in ASR for disability-adjusted life years (DALYs) from 1990 (left) to 2021 (right). (C) Trends in ASR for prevalence from 1990 (left) to 2021 (right). (D) Trends in ASR for incidence from 1990 (left) to 2021 (right). ASR = age-standardized rate, DALY = disability-adjusted life year, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

In 1990, the highest ASDR were concentrated in Western Africa, Eastern Sub-Saharan Africa, and Mongolia. Lower rates were seen in the United States, Canada, Australia, and Russia.

By 2021, the global landscape had changed. High-income countries including the United States, Canada, and Australia experienced a clear increase in ASDR. Similarly, the age-standardized prevalence and incidence rates showed marked rises in North America and Australia over the 3 decades, indicating a broadening geographic impact of the disease.

4. Sociodemographic index (SDI) stratified analysis

The analysis of trends from 1990 to 2021, stratified by global sociodemographic index (SDI) quintiles, revealed distinct epidemiological patterns for EO-NRLC in terms of age-standardized prevalence (ASPR), incidence (ASIR), mortality (ASMR), and disability-adjusted life year (DALY) rates. Future projections based on SDI modeling further highlight regional disparities in the disease trajectory (Fig. 3A–D; Supplementary Materials 1–4, Supplemental Digital Content, https://links.lww.com/MD/R492).

Figure 3.

EO-NRLC age-standardized rates (stratified by sociodemographic index, SDI) globally and across the 21 GBD regions from 1990 to 2021: (A) mortality, (B) disability-adjusted life years (DALYs), (C) prevalence, and (D) incidence. EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH), GDB = global burden of disease, SDI = sociodemographic index.

4.1. Regional trends and SDI association

Most regions exhibited an overall increase in disease burden metrics over time. However, notable nonlinear patterns emerged in specific areas:

In Southern Sub-Saharan Africa, all ASRs rose sharply with increasing SDI until ~0.6, then declined, before increasing again as SDI decreased further.

In the High-income Asia Pacific region, rates increased progressively, peaked around an SDI of 0.8, and then showed a sharp decline followed by a gradual recovery.

In Australasia, the increases in ASIR and ASPR associated with rising SDI were more pronounced than in other regions.

Correlation analysis indicated a significant negative relationship between SDI and ASMR (r = −0.2828, P < .001), ASDR (r = −0.3309, P < .001), and ASIR (r = −0.1795, P < .001), while the correlation with ASPR was not significant (P = .511).

4.2. Gender disparities across SDI strata

A comparison of ASRs between 1990 and 2021 across the 5 SDI regions showed an overall increase in burden for both sexes. However, a clear reversal in gender disparity was observed across the development spectrum (Fig. 4):

Figure 4.

Comparison of global and SDI regional ASR and gender differences between 1990 and 2021. (A) 1990-mortality, (B) 2021-mortality, (C)1990-disability-adjusted life years (DALYs), (D) 2021-disability-adjusted life years (DALYs), (E) 1990-prevalence, (F) 2021-prevalence, (G),1990-incidence, and (H) 2021-incidence. ASR = age-standardized rate, DALY = disability-adjusted life year, SDI = sociodemographic index.

In low-SDI regions, women consistently showed higher incidence, prevalence, mortality, and DALY rates compared to men.

This pattern reversed in high-SDI regions, where the disease burden was greater among men. This finding underscores the interaction between socioeconomic development and gendered risk profiles for EO-NRLC.

5. Burden attributed to EO-NRLC by age and sex

The mortality and DALYs attributable to early-onset NASH-related liver cancer (EO-NRLC) demonstrated a clear age-dependent increase in both genders (Fig. 5).

Figure 5.

EO-NRLC burden by age and sex. (A) Mortality rates across different age groups and sexes. (B) DALYs by age and sex distribution. DALY = disability-adjusted life year, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

In 2021, among individuals under 30 years old, females consistently exhibited higher death counts and DALYs across all 3 younger age subgroups compared to males. However, a pivotal shift occurred in the 30 to 34-year age group, where the burden distribution began to transition. From age 35 onward, males progressively surpassed females in DALYs, establishing a reversal in the gender disparity that characterized younger populations. This pattern highlights a critical inflection point in the burden of EO-NRLC across the lifespan.

6. Risk factor attribution analysis

An in-depth analysis of 6 established risk factors for liver cancer was conducted across age cohorts of early-onset NASH-related liver cancer (EO-NRLC). The detailed proportion of DALYs attributable to each factor is presented in Supplementary File 5, Supplemental Digital Content, https://links.lww.com/MD/R492, with a summary shown in Figure 6.

Figure 6.

Ratio of global and regional risk factor contributions to EO-NRLC in 2021 (A) age-standardized mortality rates (B) annual standardized DALYs rates. DALY = disability-adjusted life year, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

The analysis revealed that the attributable risk from NASH itself for EO-NRLC was negligible across all age groups. In contrast, the influence of the other 5 risk factors on the ASDR progressively increased with age.

A notable transition was observed around age 30. Behavioral risks, which showed zero attribution in the 15 to 29 age group, began to rise after age 30. By age 40, behavioral risks emerged as the predominant contributing factor to the EO-NRLC burden. Among the remaining 4 risk factors, high fasting plasma glucose and broader metabolic risks demonstrated lower attribution ratios compared to smoking and tobacco use across the studied age range.

7. Decomposition analysis of burden drivers

A decomposition analysis was conducted to quantify the contributions of 3 key drivers – population aging, population growth, and changes in epidemiological rates – to the observed trends in EO-NRLC deaths, DALYs, prevalence, and incidence from 1990 to 2021 (Fig. 7; detailed data in Supplementary File 6, Supplemental Digital Content, https://links.lww.com/MD/R492).

Figure 7.

Visual decomposition analysis results are presented. (A) Decomposition of the age-standardized mortality rate (ASMR); (B) decomposition of the age-standardized DALY rate (ASDR); (C) decomposition analysis of the age-standardized prevalence (ASPR); (D) decomposition analysis of the age-standardized incidence rates (ASIR). Black dots represent the overall change in disease burden attributable to population aging, epidemiological transitions, and population growth. For each contributing factor, positive values signify an increase in EO-NRLC disease burden associated with that factor, while negative values indicate a reduction. ASDR = age-standardized DALY rate, ASIR = age-standardized incidence rate, ASMR = age-standardized mortality rate, ASPR = age-standardized prevalence, DALY = disability-adjusted life year, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

Globally, population growth was the primary driver, accounting for 65% of the increase in EO-NRLC burden, followed by population aging (28%) and epidemiological changes (7%).

Regionally, the influence of these factors varied:

Population growth exerted a strong positive effect on burden increases in most regions, particularly in East Asia and high- and high-middle-SDI areas. Conversely, it had a negative impact in high-income Asia Pacific and Central Europe.

Population aging significantly increased mortality and DALYs in regions including Central Europe, Central Sub-Saharan Africa, and High-income Asia Pacific.

Epidemiological changes positively influenced burden indicators in many affluent regions (Central Europe, High-income Asia Pacific, Western Europe, High-income North America). However, their impact was negative in several regions, including parts of East Asia, Sub-Saharan Africa, and lower SDI strata. Overall, epidemiological changes contributed more negatively to burden trends than the other 2 drivers in affected regions.

This analysis clarifies that while broad demographic forces are major contributors, region-specific changes in underlying risk or survival are also critical in shaping the EO-NRLC burden.

8. Frontier analysis for intervention potential

A frontier analysis was performed using 1990 to 2021 data to compare ASRs of mortality, DALYs, incidence, and prevalence with corresponding sociodemographic index (SDI) values (Fig. 8; detailed data in Supplementary Files 7–10, Supplemental Digital Content, https://links.lww.com/MD/R492). This aimed to identify countries with the greatest potential for improving EO-NRLC outcomes relative to their level of development.

Figure 8.

Conducting cutting-edge analysis to identify gaps in the improvement of EO-NRLC burden. (A) A frontier analysis of age-standardized mortality rates (ASMR); (B) a frontier analysis of age-standardized disability-adjusted life-year rates (ASDR); (C) a frontier analysis of age-standardized prevalence rates (ASPR); and (D) a frontier analysis of age-standardized incidence rates (ASIR). The boundary is delineated in solid black, while countries and regions are represented by dots. The top 15 countries with the largest disparities in effectiveness (those exhibiting the greatest differences in DALYs due to breast cancer at the border) are highlighted in black. Frontier countries with low SDI (<0.5) and minimal effectiveness differentials are indicated in blue, whereas examples of countries and regions with high SDI (>0.85) and relatively significant effectiveness differentials are marked in red. Red dots signify the increase in ASR for EO-NRLC from 1990 to 2021, while blue dots denote a decrease in ASR for EO-NRLC during the same period. ASDR = age-standardized DALY rate, ASIR = age-standardized incidence rate, ASMR = age-standardized mortality rate, ASPR = age-standardized prevalence, ASR = age-standardized rate, DALY = disability-adjusted life year, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

8.1. Countries with largest improvement gaps

The analysis identified countries where current outcomes lag furthest behind the expected frontier for their SDI level, signaling high potential for targeted intervention. The ten countries with the largest gaps across multiple ASR metrics included: Mongolia, Gambia, Mauritania, Eswatini, Mali, Tonga, Egypt(low to middle SDI).

9. Qatar (high SDI)

9.1. Countries with smallest improvement gaps

Conversely, several countries were found to have outcomes close to the expected frontier for their SDI. The 10 countries with the smallest gaps included: Morocco, Somalia, Mauritius, Argentina, Ukraine, Haiti, Sri Lanka, Saint Lucia, Brazil.

Most of these are low- or middle-SDI countries, though exceptions exist (Somalia, Haiti with very low SDI; Qatar with high SDI).

9.2. SDI and improvement potential

Overall, higher-SDI countries generally showed smaller gaps between observed and frontier performance, indicating more optimized outcomes relative to resources. In contrast, lower-SDI countries exhibited wider variation – some (Somalia, Haiti) performed near their frontier despite limited resources, while others (Mongolia, Gambia) showed substantial room for improvement.

These findings highlight that developmental status alone does not determine outcomes; within similar SDI ranges, substantial differences in efficiency and potential exist, which can guide priority-setting for health system interventions.

10. Age-period-cohort analysis

Mortality, DALYs, prevalence, incidence, and population data were organized into consecutive 5-year intervals from 1992–1996 (median 1994) to 2017–2021 (median 2019) and stratified by 7 successive birth cohorts, ranging from 1943–1948 (median 1945) to 2000–2004 (median 2002). Detailed values are provided in Supplementary Files 11 to 12, Supplemental Digital Content, https://links.lww.com/MD/R492.

10.1. Age-specific trends over time

All 4 EO-NRLC burden indicators increased consistently with age in both the earliest (1992–1996) and latest (2017–2021) periods (Fig. 9A–D). However, the rate of age-related increase varied across time:

Figure 9.

Impact of age, period, and cohort on global mortality, DALYs rates, prevalence, and incidence. (A) EO-NRLC cohort-specific mortality rates by time period. (B) EO-NRLC cohort-specific mortality rates by age group. (C) EO-NRLC cohort-specific DALYs rates by period. (D) EO-NRLC cohort-specific DALYs rates by age group. (E) EO-NRLC cohort-specific Prevalence by period. (F) EO-NRLC cohort-specific Prevalence by age group. (G) EO-NRLC cohort-specific Incidence by time period. (H) EO-NRLC cohort-specific Incidence by age group. (I) Local and net drift values of the global EO-NRLC prevalence from 1992 to 2021. (J) Fitted longitudinal age curves for the prevalence (per 1,00,000 person-years) along with the corresponding 95% confidence intervals. (K) The relative risk of prevalence in each period was estimated while accounting for age and nonlinear cohort effects. The corresponding 95% confidence intervals were also compared against the reference periods of 1992 to 1996 and 2002 to 2006. (L) Relative risks of patient incidence rates (for the cohort 1955–1959) were adjusted for age and nonlinear period effects. Corresponding 95% confidence intervals were calculated for each cohort when compared to the reference cohorts (1967–1972 and 1982–1987). DALY = disability-adjusted life year, EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

Mortality and DALYs rose most rapidly with age during 1997 to 2001, and most slowly during 2002 to 2010.

Prevalence increased most sharply with age in 2012 to 2016, and most gradually in 1992 to 1996.

Incidence showed its steepest age-related rise in 2012 to 2016, and its slowest in 2002 to 2006.

Notably, while most age groups exhibited gradual upward trends across birth cohorts, pronounced fluctuations were observed in the 40 to 44 and 45 to 49 year age groups.

10.2. Drift, age, period, and cohort effects

Net drift (overall annual change) and local drift (age-specific annual change) were estimated from rates and population data (Fig. 9E–H):

Local drift (Fig. 9E): For ages 25 to 49, local drift remained below average and relatively stable. In contrast, ages 15 to 19, 20 to 24, and 44 to 49 experienced above-average local drift, indicating these groups were disproportionately affected in terms of prevalence trends. Prevalence also rose gradually among individuals aged 45 and older.

Age effect (Fig. 9F): As expected, the risk of EO-NRLC increased exponentially with age.

Period effect (Fig. 9G): The period relative risk (RR) remained above 1 in all intervals except 1992 to 1996 and 2002 to 2006, showing a consistent upward trend, particularly after 2006.

Cohort effect (Fig. 9H): Incidence worsened in cohorts up to 1996, then gradually declined. Prevalence increased annually in cohorts born before 1970, declined in the 1970 to 1980 birth cohort, and has risen consistently in cohorts born after 1980, with RR remaining above 1.

This analysis underscores that recent birth cohorts face a growing burden of EO-NRLC, independent of aging and period trends, highlighting a need for targeted early-life prevention.

11. Projection of future burden (2022–2040)

Using a Bayesian APC model, we projected ASRs for EO-NRLC incidence (ASIR), prevalence (ASPR), mortality (ASMR), and DALYs from 2022 to 2040, overall and by gender (Fig. 10; detailed data in Supplementary Files 13–16, Supplemental Digital Content, https://links.lww.com/MD/R492).

Figure 10.

Changing trends and projected rates of the global EO-NRLC burden from 2022 to 2040. (A) Age-standardized mortality rate (ASMR) from 1990 to 2040; (B) age-standardized DALY rate (ASDR) from 1990 to 2040; (C) age-standardized prevalence rate (ASPR) from 1990 to 2040; (D) age-standardized incidence rate (ASIR) from 1990 to 2040; (E) ASMR for men from 1990 to 2040; (F) ASDR for men from 1990 to 2040; (G) ASPR for men from 1990 to 2040; (H) ASIR for men from 1990 to 2040; (I) ASMR for women from 1990 to 2040; (J) ASDR for women from 1990 to 2040; (K) ASPR for women from 1990 to 2040; (L) ASIR for women from 1990 to 2040. ASDR = age-standardized DALY rate, ASIR = age-standardized incidence rate, ASMR = age-standardized mortality rate, ASPR = age-standardized prevalence rate, DALY = disability-adjusted life year. EO-NRLC = early-onset nonalcoholic fatty liver cancer due to nonalcoholic steatohepatitis (NASH).

By 2040, the projected burden per 100,000 population is as follows:

Mortality (ASMR): 0.098 overall (0.100 males, 0.091 females).

DALYs (ASDR): 4.49 overall (4.71 males, 4.90 females).

Prevalence (ASPR): 0.21 overall (0.22 males, 0.20 females).

Incidence (ASIR): 0.12 overall (0.13 males, 0.11 females).

The projections indicate a continued upward trend in all 4 metrics for both sexes over the next 2 decades, with slightly higher incidence and prevalence rates among males. These findings suggest that the age-standardized burden of EO-NRLC is expected to rise persistently, underscoring the need for reinforced prevention and early intervention strategies in the coming years.

12. Discussion

12.1. Global trends and temporal evolution of EO-NRLC burden

Using data from the Global Burden of Disease database, this study demonstrates a sustained global increase in the incidence, mortality, and DALYs of early-onset EO-NRLC from 1990 to 2021. Similar long-term upward trends have been reported in previous global assessments of liver cancer and metabolic liver disease burden. The increasing contribution of EO-NRLC among individuals aged 15 to 49 years is particularly concerning, as it implies prolonged disease duration and substantial socioeconomic impact. Population aging and expansion were identified as major contributors to this growing burden, consistent with prior epidemiological decomposition analyses. Age-related pathological changes in the liver are closely associated with the development and progression of NAFLD and HCC, particularly through mechanisms involving cellular senescence and chronic inflammation.^[14,15] In addition, the global shift toward earlier disease onset extends the exposure window for disease progression. Collectively, these findings indicate that EO-NRLC represents a rapidly emerging global health challenge requiring early preventive strategies.

12.2. Sociodemographic disparities and regional heterogeneity

Substantial regional heterogeneity in EO-NRLC burden was observed across different SDI strata. Although high-SDI regions generally exhibited lower absolute ASRs, they experienced more rapid increases in incidence and DALYs over time. In contrast, many low-SDI regions continued to display disproportionately high mortality and DALY rates, reflecting disparities in healthcare access and disease management capacity. The growing prevalence of NAFLD, now affecting nearly 38% of the global population, has played a central role in reshaping the global liver disease landscape.^[16,17] While only a minority of NAFLD cases progress to cirrhosis or HCC, the sheer size of the affected population substantially amplifies cumulative disease burden.^[18,19] Furthermore, the epidemiological transition from viral hepatitis–related liver cancer toward metabolically driven etiologies has intensified these regional disparities. These patterns highlight the need for region-specific prevention and intervention policies.

12.3. Age- and sex-specific patterns of EO-NRLC burden

Clear age- and sex-specific differences were identified in EO-NRLC epidemiology. Among individuals over 30 years of age, males consistently exhibited higher prevalence, incidence, mortality, and DALY rates than females. Conversely, females under 30 years demonstrated relatively higher disease burden, indicating a reversal of the traditional sex disparity at younger ages. This phenomenon may be partly explained by sex-specific differences in fat distribution, insulin resistance, and lipid metabolism. Previous studies have shown that the risk of NASH increases nonlinearly with sex-specific adiposity patterns and metabolic dysfunction.^[20,21] Hormonal regulation and chronic inflammatory pathways may further contribute to differential disease susceptibility between men and women. These findings underscore the importance of incorporating age- and sex-specific considerations into EO-NRLC prevention strategies.

12.4. Role of metabolic and behavioral risk factors

Behavioral and metabolic risk factors were identified as major contributors to EO-NRLC burden across all age groups. Elevated fasting plasma glucose, metabolic dysfunction, smoking, and tobacco consumption accounted for a substantial proportion of EO-NRLC-related mortality and DALYs. An overall healthy lifestyle has been shown to be inversely associated with MAFLD risk, highlighting the protective role of diet and physical activity.^[22] Diets rich in anti-inflammatory nutrients and bioactive components demonstrate inverse associations with NAFLD incidence and HCC risk.^[23,24] In contrast, high consumption of sugar-sweetened beverages and red or processed meats is positively associated with EO-NRLC risk. Sleep disruption and genetic susceptibility further amplify the risk of progression to advanced liver disease.^[25] These findings reinforce the critical importance of modifiable lifestyle factors in EO-NRLC prevention.

12.5. Epidemiological transition and disease progression mechanisms

The epidemiological landscape of liver cancer is undergoing profound changes driven by rising obesity, increased alcohol consumption, and improved management of viral hepatitis. NASH–related liver cancer has emerged as a dominant contributor to the global increase in liver cancer incidence and mortality.^[26,27] The burden of metabolic dysfunction-associated fatty liver disease varies substantially across regions, shaped by food insecurity, healthcare access, and dietary quality.^[28] NAFLD contributes the largest increase in DALYs among all liver diseases worldwide, despite lower progression rates to advanced disease.^[29] Socioeconomic disadvantage is strongly associated with increased NAFLD risk, faster disease progression, and higher mortality.^[30,31] These transitions emphasize the growing dominance of metabolic pathways in EO-NRLC pathogenesis.

12.6. Smoking, environmental exposure, and molecular mechanisms

Tobacco exposure remains a significant and modifiable risk factor for EO-NRLC. Smoking has been identified as an independent determinant in the development of NASH and its progression to HCC.^[32] Experimental studies have demonstrated that chronic exposure to tobacco-specific nitrosamines promotes hepatic inflammation, lipid accumulation, and NF-κB activation.^[33] Secondhand smoke exposure further contributes to hepatocarcinogenesis by altering gene expression and disrupting lipid homeostasis pathways.^[34] These mechanisms collectively exacerbate metabolic liver injury and malignant transformation. Therefore, comprehensive tobacco control policies may yield substantial benefits in reducing EO-NRLC burden, particularly among younger populations.

12.7. Strengths, limitations, and public health implications

This study provides a comprehensive overview of EO-NRLC burden using standardized GBD methodologies; however, several limitations should be acknowledged. Data quality and registry coverage remain suboptimal in many low- and middle-income countries, potentially leading to misclassification or underestimation of disease burden. The absence of histological and etiological stratification in the GBD framework limits more granular analyses. Additionally, reliance on cross-sectional data restricts causal inference. Despite these limitations, the findings generate valuable hypotheses and highlight critical public health priorities. Individuals aged 15 to 49 years should be encouraged to adopt healthy lifestyles, improve dietary quality, and reduce exposure to known risk factors. Policy efforts should emphasize early prevention, equitable healthcare access, and efficient resource allocation to mitigate the growing impact of EO-NRLC.

13. Conclusion

Against the backdrop of a worldwide epidemiological transition characterized by escalating incidence rates, mortality statistics, disease prevalence, and DALYs burdens across 204 geographical entities, EO-NRLC has progressively established itself as a significant etiological factor in the global burden of hepatic malignancies. Therefore, it is essential for nations to formulate strategies for the allocation of healthcare resources tailored to their specific disease burdens, while implementing effective prevention and early detection initiatives.

Acknowledgments

Thanks to the assistance of Department of Hepatobiliary and Pancreatic Surgery, Qingdao Chengyang District People’s Hospital, Qingdao City, Shandong Province, China. This study was generously supported by Jingding Medical Tech, to whom we extend our sincere gratitude. We especially thank them for providing authorization and technical support for the JD_GBDR software.

Author contributions

Conceptualization: Nan Tang, Zengyin Chen.

Data curation: Hezhen Shang, Nan Tang, Zengyin Chen.

Formal analysis: Nan Tang.

Investigation: Hezhen Shang, Nan Tang.

Methodology: Binghui Chu.

Project administration: Hezhen Shang, Binghui Chu, Xiaoxin Yu, Zengyin Chen.

Resources: Binghui Chu, Xiaoxin Yu.

Software: Xiaoxin Yu.

Supervision: Xiaoxin Yu, Zengyin Chen.

Visualization: Zengyin Chen.

Writing – original draft: Zengyin Chen.