Epidemiological Transition in Health Risks and Disease Burden Among Women of Childbearing Age in East Asia and the Pacific, 1990-2023.

Women of childbearing age (WCBA), defined as 15–49 years by WHO, represent a critical demographic cohort, projected to number approximately 1.97 billion globally by 2023, and constitute about 24.4% of the world’s total population. 1 This demographic group not only forms the foundation of societal productivity but also bears the primary responsibility for childbirth and the health of subsequent offspring. 2 Over recent decades, significant progress has been made in the health status of WCBA within the framework of global strategic goals, such as SDG 3 and SDG 5, most notably evidenced by the decline in the global maternal mortality ratio from 318 per 100,000 live births in 1990 to 191 in 2021. 3–5 Nevertheless, persistent challenges persist, including pronounced regional inequalities, social disparities, and inherent physiological vulnerabilities during childbearing years. 6 , 7 Within the global landscape, the East Asia and Pacific (EAP) region emerges as one of the most densely populated areas. 8 This region encompasses 35 countries and territories, spanning a wide spectrum of socioeconomic development from high-income economies such as Japan and Singapore to lower-middle-income nations including Papua New Guinea and Timor-Leste. Over the past three decades, the EAP region has exhibited distinct trends, including rapid urbanization, declining fertility rates, and accelerated population aging. 9 , 10 Meanwhile, rapid epidemiological transitions are occurring within a short period of time, leading to the coexistence of established infectious diseases and rapidly emerging noncommunicable diseases, thereby creating a double burden. These trends underscore the urgency and necessity of improving the health of women aged 15–49 years for broader socioeconomic development. Moreover, health disparities in the EAP region are particularly representative due to its complex socioeconomic and ethnic diversity. For instance, in advanced economies such as Japan or South Korea, disease burdens among WCBA are predominantly characterized by late-onset gynecological cancers, mental health disorders, and lifestyle-related risks, whereas rural areas in Southeast Asia are primarily affected by infectious diseases, malnutrition, and maternal disorders. 11–13 Moreover, the COVID-19 pandemic has exacerbated existing gender inequalities, presenting women with new health challenges. 14 The evidence indicates that health interventions targeting WCBA are significantly cost-effective, yielding broad and enduring positive returns in reducing maternal and overall female mortality, improving child health outcomes, and advancing gender equality. 15 , 16 Though WCBA increasingly experience substantial burdens from non-communicable diseases, mental disorders, and musculoskeletal conditions, women’s health studies remain largely focused on reproductive health topics, lacking systematic assessments of the overall disease spectrum and comprehensive burden among WCBA. 5 , 17 , 18 Furthermore, the existing evidence is primarily based on data from 2021 or earlier, which limits the accurate assessment of the impact of the COVID-19 pandemic on the WCBA, and has not systematically quantified cross-country gaps in preventable burden. 19 Based on the Global Burden of Disease (GBD) 2023 study, this study aims to conduct a detailed and comprehensive assessment of the burden of disease and injury, as well as attributable risk factors among women aged 15–49 in the EAP region. We employ the estimated annual percentage change (EAPC) to capture long-term trends and use a frontier analysis approach, with the Socio-Demographic Index (SDI) as a benchmark, to measure country performance, thereby providing an updated description of the epidemiological shift in the burden of disease and the inequalities among countries. To our knowledge, this is the first comprehensive study using GBD 2023 data (including the COVID-19 period) and an SDI-frontier analysis to examine longitudinal trends in disease burden and risk factors among women aged 15–49 in the EAP region, and to quantify cross-country disparities.

This study was conducted using the estimates data from the GBD 2023. GBD 2023 provides systematic and comparable estimates of disease burden and attributable risk across 204 countries and territories, constituting the most extensive and detailed global health database currently available. 20 The data sources encompass a wide range of inputs, including vital registration systems, surveys, population censuses, and scientific literature. We performed a systematic analysis of these publicly available datasets, with all data accessed through the GBD Results Tool ( https://vizhub.healthdata.org/gbd-results/ ). The precise extraction settings in the tool query were specified as follows: GBD Estimate (Cause of death or injury/Risk factor); Age (15–49 years); Metric (Number, Rate); Measure (DALYs); Sex (Female); Year (Annual from 1990 to 2023); and Location (East Asia and Pacific region, along with its individual sub-national components). The analysis included all 35 countries and territories within the World Bank EAP classification: American Samoa, Australia, Brunei Darussalam, Cambodia, China, Democratic People’s Republic of Korea, Fiji, Guam, Indonesia, Japan, Kiribati, Lao People’s Democratic Republic, Malaysia, Marshall Islands, Micronesia (Federated States of), Mongolia, Myanmar, Nauru, New Zealand, Northern Mariana Islands, Palau, Papua New Guinea, Philippines, Polynesia, Republic of Korea, Samoa, Singapore, Solomon Islands, Taiwan (China), Thailand, Timor-Leste, Tonga, Tuvalu, Vanuatu, and Vietnam. For descriptive, comparative, and trend analyses, we retrieved DALY numbers and age-standardized DALY rates (ASDRs), together with their corresponding 95% uncertainty intervals (UIs), for all-cause burden, major cause groups, level 3 causes, and attributable risk factors. Causes were analyzed according to the standard GBD hierarchy, including communicable, maternal, neonatal, and nutritional diseases (CMNN), non-communicable diseases (NCDs), and injuries, while risk factors were grouped into behavioral, environmental/occupational, and metabolic domains. Crucially, due to the computational architecture of the GBD framework, which utilizes standardized predictive modeling and iterative imputation to propagate estimates across sparse domains, the retrieved data constituted a fully balanced panel matrix. All 35 locations possessed complete, non-missing modeled estimates across all 33 years, 174 causes, and 66 risk factors. Any inherent variations in primary data quality or completeness are captured and represented by the width of the 95% uncertainty intervals (UIs) rather than missing data cells. GBD 2023 introduces substantial methodological improvements compared to previous versions, significantly enhancing the robustness, comparability, and policy relevance of its estimates. 20 , 21 The major developments include the integration of over 35,000 new sources for disease and injury burden estimation and over 16,000 new sources for risk factor exposure assessment. To model complex epidemiological trends, GBD 2023 implements a novel tool, DisMod-AT (Age–Time), which enables a more accurate estimation of age-specific trends over time by accounting for cohort effects. For risk assessment, the scope and diversity of exposure data have been extensively broadened. The exposure-response relationships and the theoretical minimum risk exposure level (TMREL) were re-evaluated and corrected using the most recent evidence from cohort and case-control studies. The study population was defined as WCBA (15–49 years), consistent with the standard demographic classification of the WHO. 17 To motivate the geographic focus, we first described the absolute number and population proportion of WCBA across the seven World Bank regions. As shown in Figure 1 , the EAP region accounted for the largest number of WCBA and a substantial share of the regional population. The EAP region encompasses countries with substantial variation in demographic structure, socioeconomic development, health-system capacity, and epidemiological profiles, making it an ideal setting to investigate regional disparities and development-adjusted disease burden. Figure 1 Population size and ranking of women aged 15–49 years across World Bank regions, 1990 and 2023. ( A ) The estimated population size of WCBA in 2023 across the seven World Bank regions. Bars indicate the number of WCBA (millions), and dots indicate the proportion of WCBA in the total regional population (%). ( B ) Rank order of World Bank regions by WCBA population in 1990 and 2023, with values shown in millions. Panel A compares the seven World Bank regions in 2023. Bars use the left y-axis to show the number of women of childbearing age in millions, and dots use the right y-axis to show women of childbearing age as a percentage of the total female population. East Asia and Pacific and South Asia have the largest populations of women of childbearing age. Panel B compares regional rankings and population sizes in 1990 and 2023. East Asia and Pacific remains first, increasing from 502 million to 548 million, and South Asia remains second, increasing from 262 million to 518 million. Sub-Saharan Africa rises from fourth place, at 117 million, to third place, at 311 million, overtaking Europe and Central Asia, which changes from 211 million to 213 million. Latin America and the Caribbean remains fifth, while the Middle East and North Africa rises from seventh to sixth and North America falls from sixth to seventh. Bar-and-dot chart showing the 2023 WCBA population and its share of the total female population across seven World Bank regions, together with a comparison of regional rankings in 1990 and 2023. Abbreviations : WB, world bank; WCBA, women of childbearing age; GBD, Global Burden of Disease. Population size and ranking of women aged 15–49 years across World Bank regions, 1990 and 2023. ( A ) The estimated population size of WCBA in 2023 across the seven World Bank regions. Bars indicate the number of WCBA (millions), and dots indicate the proportion of WCBA in the total regional population (%). ( B ) Rank order of World Bank regions by WCBA population in 1990 and 2023, with values shown in millions. To systematically examine health inequalities, all countries or territories were stratified by socioeconomic development index (SDI) quintiles (low, lower-middle, middle, upper-middle, high). The SDI is a composite parameter based on income, education, and fertility data, and is routinely employed to analyze disparities in disease burden driven by socioeconomic gradients. 22 Our analysis encompassed all causes and risk factors modeled in GBD 2023 that are epidemiologically applicable to the female 15–49 age cohort. To ensure systemic consistency and avoid subjective selection bias, the inclusion criteria were structurally defined based on the GBD hierarchy: we included all standard Level 3 causes to capture 174 discrete disease entities or injuries, and selected the 66 most detailed risk factors at their finest terminal level of resolution to map risk exposures with the highest specificity. The specific disease burden is categorized and analyzed according to the GBD system, which comprises three broad Level 1 categories: CMNN, NCDs, and injuries. Risk factors were also classified into three principal domains: metabolic, behavioral, and environmental/occupational risks. In the present study, representative metabolic risks included high systolic blood pressure, high fasting plasma glucose, and high body-mass index, as defined in the GBD 2023 framework. The GBD 2023 core publications and supplementary online database provide detailed case definitions and diagnostic criteria for each disease or risk. 20 In this study, maternal disorders and mental disorders were analyzed as separate GBD cause categories. Maternal disorders refer to GBD-defined conditions related to pregnancy, childbirth, and the puerperium, whereas mental disorders refer to non-fatal mental health conditions such as depressive and anxiety disorders among females aged 15–49 years. Because the study population included all women of childbearing age rather than only pregnant women, both maternal and non-maternal causes contributed to the overall burden estimates. The burden attributable to each risk-outcome pair was calculated based on the population attributable fraction (PAF), derived from the GBD risk assessment framework. The general formula is as follows: \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$$PA{F_{joasgt}} = {{\mathop \sum \nolimits_{x = l}^m R{R_{joasg}}\left(x \right){P_{jasgt}}\left(x \right) - R{R_{joasg}}\left({TMRE{L_{jas}}} \right)} \over {\mathop \sum \nolimits_{x = l}^m R{R_{joasg}}\left(x \right){P_{jasgt}}\left(x \right)}}$$\end{document} Here, \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$PA{F_{joasgt}}$\end{document} represents the population attributable fraction for cause o associated with risk factor j in age group a , sex s , location g , and year t . The term \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$R{R_{joasg}}\left(x \right)$\end{document} denotes the relative risk corresponding to exposure level x for risk factor j and cause o , whereas \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${P_{jasgt}}\left(x \right)$\end{document} indicates the proportion of the population exposed at level x within the specified age group, sex, location, and year. \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$TMRE{L_{jas}}$\end{document} refers to the theoretical minimum risk exposure level for risk factor j according to age group and sex. 20 , 23 The main summary indicator for population health was the disability-adjusted life year (DALY), which combines years of life lost due to premature mortality (YLLs) and years lived with disability (YLDs). All DALY rates were age-standardized using the GBD global standard population and are presented as ASDRs per 100,000 population to enable cross-country and cross-temporal comparisons. The ASDR was calculated as: \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$$ASDR = {{\mathop \sum \nolimits_{i = 1}^n {\gamma _i}{\omega _i}} \over {\mathop \sum \nolimits_{i = 1}^n {\omega _i}}} \times 100,000$$\end{document} where \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${\gamma _i}$\end{document} denotes the age-specific DALY rate for age group i , \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${\omega _i}$\end{document} denotes the corresponding weight for age group i in the GBD global standard population, and n denotes the number of age groups. In this study, age standardization was applied across the 5-year age strata within women aged 15–49 years. To quantify temporal trends in disease burden from 1990 to 2023, we calculated the EAPC. This was derived by fitting a linear regression model to the natural logarithm of the annual ASRs: \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$${\mathrm{ln}}\left({AS{R_t}} \right) = \alpha + \beta \times {\mathrm{yea}}{{\mathrm{r}}_t} + {\varepsilon _t}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$ASR{_{t}}$\end{document} represents the age-standardized rate in year t, α is the intercept, β denotes the annual change on the log scale, and \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${\varepsilon _t}$\end{document} is the random error term. The EAPC was then computed as \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$EAPC = 100 \times \left({{e^\beta } - 1} \right)$\end{document} . A positive EAPC with 95% confidence interval (CI) indicates an increasing trend, whereas a negative value denotes a decreasing trend. 24 In order to specifically assess the impact of the COVID-19 pandemic on the WCBA disease burden, we estimated EAPCs separately for two distinct phases: the period of rising burden (2019–2021) and the period of declining burden (2021–2023, 25 ). To benchmark disease burden against socioeconomic development, we conducted a frontier analysis relating location-year ASDRs among WCBA to the corresponding socio-demographic index (SDI). The frontier was estimated using a non-parametric bootstrap procedure with 100 iterations. In each iteration, a bootstrap sample of the same size as the original location-year dataset was drawn with replacement, ordered by ascending SDI and descending ASDR, and the frontier was defined as the cumulative minimum ASDR across the ordered observations. We then summarized the bootstrap distribution of the frontier using the mean and the 2.5th and 97.5th percentiles. A locally weighted regression (LOESS) curve with a span of 0.3 was applied to the bootstrap-derived frontier values to obtain a smooth development-specific frontier. The 100-iteration bootstrap sample serves to map frontier boundary uncertainty without propagating outlier noise, while the 0.3 span parameter was selected to balance local non-linear inflections against long-term development gradients. Distance-to-frontier was evaluated using two metrics: (1) Absolute distance, calculated as the difference between the observed ASDR and the corresponding frontier ASDR at the same SDI, where values greater than 0 indicate excess burden relative to the best observed performance; and (2) Frontier distance ratio, defined as the ratio of the frontier ASDR to the observed ASDR (Frontier ASDR/Observed ASDR), ranging from 0 to 1, where a value closer to 1 indicates that a country is approaching its optimal performance relative to its developmental stage. All estimates are reported with their corresponding 95% uncertainty intervals (UIs). Statistical significance of differences or trends was inferred when the 95% UIs did not overlap. This conservative criterion was intentionally adopted to mitigate Type I errors, as traditional sample-based hypothesis tests (p-values) tend to underestimate the mathematical uncertainty generated during the iterative GBD modeling and imputation processes. Data extraction, processing, statistical analysis, and figure preparation were completed collaboratively by the authors using R software (version 4.3.0). The complete R code for the frontier analysis is provided in Supplementary File 1 . The analytical workflow was prespecified on the basis of the GBD 2023 framework, including extraction of DALY numbers and ASDR in the EAP region and its 35 countries and territories from 1990 to 2023, as well as analyses of causes, attributable risk factors, temporal trends, and SDI-related frontier patterns. Because this study was based on publicly available GBD 2023 modeled estimates rather than newly collected individual-level data, conventional sample size or statistical power calculations were not directly applicable. We therefore focused on the uncertainty framework of GBD and reported 95% uncertainty intervals for the major estimates. To reduce analysis bias, we applied standardized GBD definitions and estimation procedures, used consistent prespecified analytical criteria, and interpreted the findings with caution in light of potential data sparsity, misclassification, uncertainty propagation, and the log-linearity assumption underlying EAPC estimation. This study is based on anonymized, aggregated secondary data from the GBD study. Ethical approval and informed consent were obtained during the original data collection processes by GBD collaborators. No individual-level data were used, and thus no additional ethical approval was required for this analysis. The reporting of this research adheres to the Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER). 26

Based on GBD 2023 modeled estimates, across 1990–2023, the overall DALY profile in WCBA in the EAP region was consistently dominated by NCDs, with a clear long-term shift in composition away from injuries and CMNN. The proportional contribution of NCDs rose from 62.29% in 1990 to 75.65% in 2023. In contrast, the share attributable to injuries decreased from 19.73% in 1990 to 11.92% in 2023, while CMNN diseases declined from 17.98% to 12.43% over the same period. Notably, the disease burden of CMNN showed a significant increase in 2021, subsequently declining ( Figure 2A and C ). Figure 2 Temporal trends in DALYs among women aged 15–49 years in EAP by cause groups and risk factor, 1990–2023. Total DALYs in EAP by major cause ( A ) and risk factor ( B ) shown as DALY numbers (bars) and DALY rates per 100,000 population (lines), 1990–2023. ( C ) Proportional composition of total DALYs (%) by major cause ( C ) and risk factor ( D ), 1990–2023. Panels A and B combine stacked bars showing DALY numbers on the left y-axes with lines showing DALY rates per 100,000 population on the right y-axes from 1990 to 2023. Panel A separates non-communicable diseases, injuries, and communicable, maternal, neonatal, and nutritional diseases. Non-communicable diseases account for the largest burden, while the DALY rates for all three cause groups decline overall, most slowly for non-communicable diseases. Panel B shows DALYs attributable to behavioral, environmental and occupational, and metabolic risk domains. Panels C and D show the annual proportional composition. In Panel C, the share of DALYs due to non-communicable diseases increases from 62.29 percent in 1990 to 75.65 percent in 2023. In Panel D, the combined proportion of DALYs attributable to the three risk domains decreases from 38.44 percent to 35.89 percent, with a smaller contribution from environmental and occupational risks and a larger contribution from metabolic risks by 2023. Four-panel figure showing trends in DALY numbers and rates and changes in proportional composition by major cause group and risk-factor domain, 1990–2023. Abbreviations : DALYs, disability-adjusted life years; WCBA, women of childbearing age; GBD, Global Burden of Disease. Temporal trends in DALYs among women aged 15–49 years in EAP by cause groups and risk factor, 1990–2023. Total DALYs in EAP by major cause ( A ) and risk factor ( B ) shown as DALY numbers (bars) and DALY rates per 100,000 population (lines), 1990–2023. ( C ) Proportional composition of total DALYs (%) by major cause ( C ) and risk factor ( D ), 1990–2023. To provide statistical support for the broad temporal patterns shown in Figure 2 , we further quantified the long-term trends in ASDRs for the three major cause categories. All three broad categories showed declining ASDRs between 1990 and 2023, with the steepest decrease observed for injuries (EAPC: −2.43, 95% CI: −2.58 to −2.28), followed by communicable, maternal, neonatal, and nutritional diseases (EAPC: −2.03, 95% CI: −2.36 to −1.70), whereas non-communicable diseases declined more slowly (EAPC: −0.30, 95% CI: −0.39 to −0.21) ( Supplementary Table S1 ). Regarding risk factors, the burden of the three major attributable risk factors showed a moderate downward trend, decreasing from 38.44% in 1990 to 35.89% in 2023. Specifically, the absolute burden of environmental/occupational risks declined more significantly, falling from 12.49% in 1990 to 8.26% in 2023, while metabolic risks increased from 7.34% to 10.66% ( Figure 2B and D ). To provide statistical support for the broad temporal patterns shown in Figure 2 , we further quantified the long-term trends in age-standardized DALY rates (ASDRs) for the three major risk-factor domains. Between 1990 and 2023, ASDRs attributable to behavioral risks and environmental/occupational risks declined, with the steepest decrease observed for environmental/occupational risks (EAPC: −2.19, 95% CI: −2.36 to −2.02). In contrast, ASDRs attributable to metabolic risks increased significantly over time (EAPC: 0.23, 95% CI: 0.14 to 0.33) ( Supplementary Table S2 ). In detail, the ranking of causes of disease burden among WCBA in the EAP region from 1990 to 2023 shifted from a mixed pattern dominated by injuries, gynecological diseases, and infectious diseases to patterns characterized by nonfatal, highly disabling diseases. In 1990, self-harm ranked first (ASDR 1169.0[800.1–1536.2] per 100,000 population), followed by road injuries and gynecological diseases. Since 1996, gynecological diseases have been the leading cause of disease burden (only surpassed by COVID-19 in 2021), while headache disorders and low back pain remained persistently high-ranking. Notably, mental disorders (such as depression and anxiety disorders) have become increasingly prominent in recent years ( Figure S1 ). Similarly, the primary risk factors for DALYs also underwent significant changes between 1990 and 2023. In 1990, the largest attributable burden was associated with occupational injuries (1198.2 [876.6–1594.4]), followed by intimate partner violence (786.0 [127.7–1494.7]) and iron deficiency (674.2 [433.1–1042.1]). Over time, cardiovascular metabolic risks steadily increased. By the mid-2000s, high systolic blood pressure became the primary risk factor. Since 2016, high body mass index (BMI) took the top position and maintained its lead through 2023 (583.5[260.1–913.5]). In 2023, the top three risk factors were all metabolic conditions: high body-mass index, high systolic blood pressure, and high fasting plasma glucose ( Figure S2 ). Figure 3 presents the most significant changes in causes of death and risk factors from 1990 to 2023. COVID-19 showed the most pronounced increase (EAPC: 562.13%), reflecting the emergence of COVID-19 from a zero baseline before 2020 rather than a conventional long-term epidemiological trend, followed by police conflict and executions (5.91%) and other neurological disorders (4.30%). In contrast, infectious diseases declined most rapidly, including Zika virus (−24.12%), measles (−9.11%), and tetanus (−7.19%) ( Figure 3A ). Figure 3 Top 10 causes ( A ) and risk factors ( B ) with the largest EAPC in DALY rates among women aged 15–49 years in East Asia and Pacific, 1990–2023. Two horizontal diverging bar charts compare the ten largest positive and ten largest negative estimated annual percentage changes in DALY rates from 1990 to 2023. Panel A shows causes. Cause-level EAPCs range from −24.12 percent for Zika virus to 562.13 percent for COVID-19, whose bar is markedly longer than all others. Other notable increases include police conflict and executions at 5.91 percent, other neurological disorders at 4.30 percent, and HIV/AIDS at 3.31 percent, while notable decreases include measles at −9.11 percent and tetanus at −7.19 percent. Panel B shows risk factors on a narrower scale. The largest increases are for a diet high in sugar-sweetened beverages at 4.56%, a diet high in red meat at 3.60 percent, and low temperature at 3.23%. The largest decreases are for unsafe sanitation at −5.37 percent, no access to a handwashing facility at −4.86 percent, and household air pollution from solid fuels at −4.30 percent. Red bars indicate increases and blue bars indicate decreases. Two horizontal bar charts showing the ten largest increases and decreases in cause- and risk-factor DALY-rate EAPCs, 1990–2023. Abbreviations : DALYs, disability-adjusted life years; WCBA, women of childbearing age; EAPC, estimated annual percentage change. Top 10 causes ( A ) and risk factors ( B ) with the largest EAPC in DALY rates among women aged 15–49 years in East Asia and Pacific, 1990–2023. For risk factors, the largest increases were concentrated in dietary exposures, with diet high in sugar-sweetened beverages (EAPC 4.56%) and diet high in red meat (3.60%) being the most significant. Conversely, the largest decreases were observed in unsafe sanitation (−5.37%), no access to handwashing facility (−4.86%), and household air pollution from solid fuels (−4.30%) ( Figure 3B ). To evaluate the impact of the COVID-19 pandemic on disease burden causes and risk factors, we compared the EAPC in ASDR between 2019–2021 (x-axis, rising phase) and 2021–2023 (y-axis, declining phase). Regarding various causes, most points cluster near the origin, indicating that the COVID-19 pandemic had a relatively mild impact on their EAPC, while some other diseases exhibit distinct pandemic-related inflection patterns. The disease burden of protein–energy malnutrition and exposure to forces of nature increased during 2019–2021, followed by a decline after 2021. In contrast, lower respiratory infections showed a decrease during the rising phase of the pandemic, rebounding during the declining phase ( Figure 4A ). Figure 4 Quadrant analysis of causes ( A ) and risk factors ( B ) changes in DALY rates before and after the acute COVID-19 period among women aged 15–49 years in East Asia and Pacific, 2019–2023. Dashed lines at 0% divide the plot into four quadrants representing (increase–increase), (increase–decrease), (decrease–increase), and (decrease–decrease) patterns across the two periods. The right arrows (→) in the legend and labels denote the sequential transition or direction of change from the initial period (2019–2021) to the subsequent period (2021–2023). Bubble size is proportional to the burden in 2021 (DALY rate per 100,000). Two bubble plots compare EAPCs in DALY rates during 2019 to 2021 on the x-axis with those during 2021 to 2023 on the y-axis. Dashed zero lines divide each plot into decrease–decrease, decrease–increase, increase–decrease, and increase–increase quadrants. Bubble size represents the DALY rate in 2021. Panel A displays causes. Lower respiratory infections change from a decrease in 2019 to 2021 to an increase in 2021 to 2023; anxiety disorders and decubitus ulcer increase in both periods; exposure to forces of nature and protein–energy malnutrition increase and then decrease; and trachoma and leishmaniasis decrease in both periods. Panel B displays risk factors. Unsafe sanitation and no access to a handwashing facility change from decreases to increases; occupational exposure to asbestos and a diet high in red meat increase in both periods; child underweight and child wasting increase and then decrease; and iron deficiency and household air pollution from solid fuels decrease in both periods. Two bubble plots comparing cause- and risk-factor DALY-rate EAPCs in 2019–2021 and 2021–2023. Abbreviations : DALYs, disability-adjusted life years; WCBA, women of childbearing age; EAPC, estimated annual percentage change. Quadrant analysis of causes ( A ) and risk factors ( B ) changes in DALY rates before and after the acute COVID-19 period among women aged 15–49 years in East Asia and Pacific, 2019–2023. Dashed lines at 0% divide the plot into four quadrants representing (increase–increase), (increase–decrease), (decrease–increase), and (decrease–decrease) patterns across the two periods. The right arrows (→) in the legend and labels denote the sequential transition or direction of change from the initial period (2019–2021) to the subsequent period (2021–2023). Bubble size is proportional to the burden in 2021 (DALY rate per 100,000). Regarding risk factors, the disease burden attributable to child wasting and child underweight increased between 2019 and 2021, and declined in the pandemic declining phase. In contrast, risks associated with water, sanitation, and hygiene (WASH), such as unsafe sanitation and no access to handwashing facility, showed a downward trend during the rising phase of the pandemic but worsened during the declining phase ( Figure 4B ). In 2023, the disease burden causes and contributing risk factors of the WCBA exhibited significant variations across countries. At the regional level, gynecological disorders ranked highest, followed by depressive disorders and headache disorders. The primary causes of disease burden in high-SDI countries were non-fatal musculoskeletal and mental health conditions, such as low back pain, ranking first in Japan, Singapore, and South Korea, and anxiety disorders ranking first in Australia and New Zealand. In contrast, low-SDI countries continue to face significant burdens from infectious and maternal disorders, including malaria, ranking first in Papua New Guinea and second in Solomon Islands, and maternal disorders ranking first in Kiribati and Timor-Leste ( Figure 5A ). Figure 5 Cross-country heterogeneity in the top 10 causes ( A ) and risk factors ( B ) for DALYs among women aged 15–49 years in East Asia and Pacific, 2023. Locations are ordered from highest to lowest SDI in 2023 (top to bottom). Two rank heatmaps show the ten leading causes and risk factors for DALYs in each East Asia and Pacific location in 2023. Locations are shown as rows and are ordered from the highest to the lowest SDI, while columns represent ranks 1 through 10. In Panel A, each cell contains a cause code from C1 to C31; in Panel B, each cell contains a risk-factor code from R1 to R21. Cell colors identify the corresponding cause or risk-factor code in the legends and do not represent burden magnitude. Gynecological disorders, depressive disorders, and headache disorders recur among the leading causes across many locations, while lower respiratory infections and maternal disorders appear among the leading causes in several lower-SDI locations. High body-mass index, high systolic blood pressure, and high fasting plasma glucose recur among the leading risk factors. Unsafe sex is prominent in several Southeast Asian locations, while sexual violence against children appears among the leading risks in several high-SDI locations. Two heatmaps showing the ten highest-ranked causes and risk factors in 2023 for each EAP location, with locations ordered by SDI. Abbreviations : DALY, disability-adjusted life year; SDI, Socio-demographic Index. Cross-country heterogeneity in the top 10 causes ( A ) and risk factors ( B ) for DALYs among women aged 15–49 years in East Asia and Pacific, 2023. Locations are ordered from highest to lowest SDI in 2023 (top to bottom). A similarly heterogeneous pattern was observed for risk factors. Cardiovascular and metabolic risks are generally severe, with high body mass index, hypertension, and high systolic blood pressure repeatedly appearing among the primary risk factors in many countries and regions. Unsafe sex is particularly prominent in Southeast Asian countries such as Thailand, Vietnam, Myanmar, Cambodia, and Timor-Leste. Meanwhile, sexual violence against children is especially severe in some high-SDI countries, including South Korea, Japan, New Zealand, and Australia ( Figure 5B ). Across the EAP region, all-cause disease burden among WCBA showed a clear age gradient over the entire period, showing a significant increasing trend from ages 15–19 to 45–49 years. Over time, age-specific DALY rates generally declined in all age groups, but experienced a noticeable short-term perturbation around 2020–2022 ( Figure 6A ). Figure 6 Temporal trends and SDI frontier of all-cause ASDR among women aged 15–49 years, with distance-to-frontier for EAP, 1990–2023. ( A ) Temporal trends in all-cause age-specific DALY rates among women aged 15–49 years in EAP stratified by 5-year age group, 1990–2023. ( B ) Temporal trends in all-cause ASDR among women aged 15–49 years across EAP countries and territories, 1990–2023. ( C ) SDI frontier analysis of all-cause ASDR. Points represent location-year observations, and the frontier curve represents the lowest expected ASDR achievable at each SDI level. ( D ) Distance to frontier in 2023 for EAP countries and territories, defined as the gap between observed ASDR and the frontier ASDR at the corresponding SDI. Panel A shows age-specific DALY rates for seven 5-year age groups from 1990 to 2023. Rates are higher in older age groups and generally decline over time, with temporary increases around 2020 to 2022. Panel B shows all-cause age-standardized DALY rates for 35 East Asia and Pacific countries and territories. The country-level rates vary substantially; in 2023, they range from approximately 13,400 per 100,000 population in Singapore to 37,647 per 100,000 population in Nauru, and many country series show a temporary increase around 2021. Panel C plots location-year ASDR observations against SDI, with color indicating year. The frontier curve represents the lowest expected ASDR achievable at each SDI level. Because the ASDR axis is reversed, lower disease burden appears higher on the plot. Panel D shows distance-to-frontier measures in 2023. Colored bars use the left y-axis to show the absolute difference between the observed and frontier ASDRs, while the black line uses the right y-axis to show the frontier efficiency ratio on a scale from 0 to 1. Pacific Island locations such as Nauru and Solomon Islands have relatively large absolute gaps, whereas China and Singapore lie close to the frontier. Four-panel image showing age-specific DALY-rate trends, country-level all-cause ASDR trends, the SDI frontier, and 2023 distance-to-frontier measures in EAP. Abbreviations : ASDR, age-standardized DALY rate; DALY, disability-adjusted life year; SDI, socio-demographic index; EAP, East Asia and Pacific. Temporal trends and SDI frontier of all-cause ASDR among women aged 15–49 years, with distance-to-frontier for EAP, 1990–2023. ( A ) Temporal trends in all-cause age-specific DALY rates among women aged 15–49 years in EAP stratified by 5-year age group, 1990–2023. ( B ) Temporal trends in all-cause ASDR among women aged 15–49 years across EAP countries and territories, 1990–2023. ( C ) SDI frontier analysis of all-cause ASDR. Points represent location-year observations, and the frontier curve represents the lowest expected ASDR achievable at each SDI level. ( D ) Distance to frontier in 2023 for EAP countries and territories, defined as the gap between observed ASDR and the frontier ASDR at the corresponding SDI. Significant disparities exist in disease burden levels and temporal trends across countries and regions. In 2023, the ASDR ranged from 13,400 per 100,000 population (Singapore) to 37,647 per 100,000 population (Nauru). Some regions experienced transient mortality surges, such as Myanmar in 2008, while others demonstrated relatively stable improvement trends ( Figure 6B ). The pandemic has significantly impacted disease burdens across countries and regions, with 33 out of 35 EAP region countries and regions experiencing higher ASDR in 2021, except for China and the Democratic People’s Republic of Korea. To further explore disease burdens across countries and regions at different development levels within the EAP region, frontier analysis compared ASDR with the SDI ( Figure 6C and D ). The distance from the frontier in 2023 highlights significant disparities in unrealized potential for burden reduction across countries and regions. Some Pacific Island countries and regions exhibit the greatest distance from the frontier, such as Nauru and Solomon Islands. In contrast, China and Singapore approach frontier levels, indicating relatively efficient disease burden performance relative to corresponding SDI levels.

Based on GBD 2023 modeled estimates, WCBA in EAP experienced a marked epidemiological transition from 1990 to 2023, with disease burden shifting from maternal and infectious conditions toward non-communicable, metabolic, mental health, and disability-related disorders. High BMI emerged as the leading attributable risk factor. Frontier analysis demonstrated an approximately 2.8-fold difference in avoidable burden between countries with similar SDI levels, indicating substantial unrealized potential for burden reduction. These burden patterns support context-specific health strategies, including strengthened mental health and chronic disease management in high-SDI settings and continued investment in maternal, infectious disease, and metabolic risk prevention in lower-SDI settings. Because these findings are derived from modeled GBD estimates, they should be interpreted descriptively rather than causally. Nevertheless, they provide updated evidence to support integrated life-course approaches to women’s health in the region.

The disease burden trajectory observed among WCBA in EAP from 1990 to 2023 conforms closely to the core tenets of epidemiological transition model, 27 characterized by a systematic displacement of communicable, maternal, neonatal, and nutritional (CMNN) conditions by non-communicable diseases (NCDs) as the dominant source of health loss. Based on the GBD 2023 modeled estimates, our findings suggest that the epidemiological transition in the EAP region has been particularly compressed over the past three decades. Although these patterns are directly supported by the GBD estimates, the mechanisms underlying these transitions cannot be inferred from the present descriptive analysis. These shifts are characterized by a persistent decline in the disease burden attributable to traditional causes, such as infectious diseases and maternal conditions, while the NCD share of total DALYs rose from 62.3% to 75.7% over just three decades. This pattern is consistent with the updated perspective that women’s health has progressively moved beyond reproduction and now requires a more comprehensive and holistic viewpoint. 28 , 29 Additionally, the EAP region exhibits significant intra-regional disparities, with countries and regions at similar SDI levels showing markedly different gaps. This suggests a substantial portion of the burden could potentially be alleviated through improved prevention, early detection, and targeted policies. A crucial perspective for interpreting the decline in maternal burden is temporally associated with policy prioritization and health-system strengthening, and improved access to essential services. The 1994 International Conference on Population and Development (ICPD) redefined reproductive health as a right and prioritized access to family planning and maternal health services. 30 The 1995 Beijing Platform for Action further institutionalized women’s health and promoted equitable access to quality services. 31 Subsequent global policies strengthened financing and responsibility for maternal and reproductive health. These include the Every Woman Every Child initiative launched by the United Nations in 2010 to advance the health and well-being of women and children, 32 the WHO’s Sustainable Development Goal (SDG) 3.1 to reduce maternal mortality 33 and 3.4 34 to improve reproductive health in 2015. Consistent with these policy trajectories, global maternal mortality has declined substantially since 2000, though progress remains unequal and unstable. 35 , 36 This situation closely aligns with the persistent maternal disorders burden and cross-national disparities observed in our study. Concurrently, the growing burden of noncommunicable diseases among WCBA should be recognized as an expected outcome of demographic shifts, lifestyle changes, and urbanization. 37 Contemporary frameworks emphasize that NCD risks accumulate across the female life course, with sex-specific exposures (including pregnancy-related complications) shaping long-term cardiometabolic vulnerability. 38 , 39 Consistent with this, our risk-factor results indicate that metabolic risks have become increasingly prominent in EAP WCBA, with rapid increases in high BMI (ASDR 583.5 per 100,000) and diet-related risks. Although the present analysis cannot determine the biological mechanisms underlying these changes, several explanations proposed in previous studies may help interpret the observed rising in metabolic and gynecological disease burden among WCBA in EAP. First, previous studies have suggested that declining fertility rates and delayed childbearing may increase cumulative lifetime estrogen exposure, 40 which has been implicated in the pathogenesis of endometriosis, uterine fibroids, and certain gynecological malignancies. 41 Second, rapid urbanization and the accompanying nutrition transition may contribute a marked increase in high BMI, a well-established risk factor not only for cardiovascular and metabolic diseases but also for gynecological conditions including polycystic ovary syndrome and endometrial cancer. 42 Third, the rising prevalence of metabolic syndrome components provides a shared biological substrate linking dietary and obesity-related risks to both cardiovascular and reproductive health outcomes, particularly insulin resistance and chronic low-grade inflammation. 43 , 44 These mechanistic intersections underscore the inadequacy of siloed care models and support the integration of metabolic risk screening into routine reproductive health services. Our GBD-based estimates suggest that the pronounced intra-regional heterogeneity revealed underscores the necessity for stratified and targeted policy responses. These findings may have several policy implications. For high-SDI settings such as Japan and Singapore, a significant portion of disease burden among WCBA is attributable to nonfatal, but highly disabling diseases, like mental disorders and chronic pain syndromes, particularly headaches and low back pain, are increasingly becoming major contributors. Global evidence also indicates that since 1990, mental disorders have consistently ranked among the leading causes of health loss, while headaches disproportionately affect WCBA. 45 , 46 Furthermore, despite international guidelines such as the United Nations Declaration on the Elimination of Violence against Women (1993), modeled estimates indicate intimate partner violence (IPV) remains prevalent and has profound impacts on physical and mental health in WCBA. 47 Considering the coexistence of communicable diseases and NCDs, the persistent burden of maternal disorders, infectious diseases and unsafe sex remains a salient challenge in low-SDI and pacific island nations, such as Papua New Guinea and the Solomon Islands, where HIV epidemics are increasingly concentrated in key populations. 48 The disease burden and underlying causes reveal the current gaps and inequalities in legal protections, education, and economic security faced by WCBA, and underscores the necessity for tiered and country-specific health strategies. 49 At the regional level, the risk-factor in EAP represents a unique epidemiological characteristic, marked by pronounced risk shifts. First, the marked decline in risks related to unsafe water, poor sanitation, and inadequate hygiene is consistent with sustained improvements in WASH coverage during the Millennium Development Goal (MDG) era and the subsequent SDG period. 50 , 51 However, these advancements are increasingly counterbalanced by rising commercial and metabolic risks, including higher intake of sugar-sweetened beverages and processed foods, alongside the increasing prevalence of high BMI. Global population-based evidence shows that obesity has increased widely since 1990, with particularly pronounced levels and trajectories in EAP regions, such as Polynesia and Micronesia. 52 Environmental and urban exposures represent another emerging risk. We observe an upward trend in environmental particulate matter pollution, consistent with external evidence indicating that PM2.5 remains a major global health risk, with persistent exposure issues in many Asian megacities. 53 China’s clean-air actions provide a notable example demonstrating that strict regulation can rapidly produce reductions in PM2.5 concentrations and yield measurable public health benefits. 54 , 55 Yet, our results suggest that ambient air pollution continues to pose an increasingly important risk for WCBA in EAP, suggesting that challenges such as rising traffic volumes, industrial emissions, and increased biomass burning emissions remain serious in EAP. 56 The COVID-19 pandemic was not only one of the most significant public health events in the past two decades, 57 but was also associated with substantial disruption to other diseases and risk factors; DALYs increased in 33 of 35 EAP locations during 2019–2021. The increase in protein-energy malnutrition, accompanied by a concurrent rise in nutrition-related risks (child wasting and child underweight), is associated with pandemic-induced income shocks, disruptions to food systems, and interruptions in routine health and nutrition services. 58 , 59 The subsequent decline after 2021 aligns with economic reopening and partial restoration of service delivery. In contrast, lower respiratory infections decreased between 2019 and 2021 but rebounded from 2021 to 2023, highlighting the critical role of non-pharmaceutical interventions and the resurgence following the easing of restrictions. 60 , 61 This underscores the necessity of establishing resilient primary healthcare and public health systems to ensure the continuity of essential nutrition and healthcare services during emergencies. At the global level, the EAP region accounts for nearly one-third of the world’s WCBA, our findings contribute to the evolving discourse on women’s health beyond reproduction. The long-term impact of reproductive events on the health of this specific population deserves careful attention. The observation that gynecological disorders and headaches are the leading causes of DALYs in 2023, yet are often neglected in global health financing, calls for a recalibration of the SDG indicators for women’s health. This conclusion aligns with the international effort to integrate NCDs prevention into reproductive and maternal health programs. 62 Concurrently, the pronounced mental health issues and chronic pain-related disease burden underscore the need to specifically integrate mental health and functional status indicators when measuring women’s health outcomes, while strengthening primary care management pathways for common high-burden syndromes, such as low back pain. 63 Finally, public health emergencies like the COVID-19 pandemic underscore the importance of building resilient health systems. 64 Future efforts must prioritize ensuring service continuity for reproductive health and chronic disease management during crises, guaranteeing this critical population access to essential, sustained medical care under any circumstances. This study leverages a standardized, comparable framework to characterize long-term trends and cross-location heterogeneity in disease burden and risk factors among WCBA in the EAP region. First, by leveraging GBD 2023 data with its enhanced methodological apparatus (including DisMod-AT and updated TMREL estimates), we provide the most current and methodologically rigorous assessment of disease burden and risk factor transitions among WCBA in EAP, extending temporal coverage through the post-pandemic period to 2023. Second, the application of phase-specific EAPC analysis (2019–2021 vs 2021–2023) offers novel quantitative evidence on the magnitude and direction of pandemic-related perturbations in this population. Third, the frontier analysis quantifies the development-adjusted avoidable burden, demonstrating up to a 2.8-fold gap between the best and worst countries. These findings carry stratified policy implications. In high-SDI settings, where mental disorders and chronic pain now dominate, health systems should integrate mental health and chronic pain management into primary care platforms. For rapidly transitioning middle-SDI countries facing the steepest metabolic risk escalation, fiscal measures targeting dietary risks should be coupled with opportunistic NCD screening embedded within reproductive health services. In low-SDI and Pacific Island settings, completing the maternal and infectious disease agenda remains paramount, yet low-cost metabolic surveillance should be established in parallel, and the COVID-19 experience underscores the critical need for health system resilience to safeguard essential services during external shocks. However, because the GBD study relies on statistical modeling to generate comparable estimates across countries and years, particularly where primary health data are sparse, cross-country comparisons should also be interpreted with caution because uncertainty may vary according to differences in data availability, surveillance systems, and model inputs across locations. Therefore, our findings should be interpreted as modeled estimates rather than directly observed epidemiological data. Second, the extremely high EAPC observed for COVID-19 should be interpreted in the context of its absence before 2020 and therefore does not represent a conventional temporal trend, the attribution of burden for rapidly emerging causes remains subject to evolving case definitions and reporting practices. Third, frontier analysis is comparative rather than causal, meaning that smaller gaps do not prove policy optimality. Nevertheless, as a performance indicator, it can help prioritize more in-depth country diagnostics. Finally, EAPC-based trend summaries assume log-linear change within each period and should be interpreted as average temporal signals rather than short-term causal effects.