Spatial-temporal dynamics of HZ prevalence among high-risk groups: a global multiregional cohort analysis since 1990 Authors: Yuan Yin, Zhao Zhe, Sai Zhang, Chengyuan Dong, Qun Lu, Chaoyan Yue Published: 2025 ABSTRACT BackgroundHerpes zoster (HZ) is caused by the reactivation of the varicella-zoster virus (VZV). Individuals aged 50 years and older are at high risk due to immune aging. The Global Burden of Disease Study (GBD 2021) provides multidimensional data across regions and time series for analyzing HZ epidemiological characteristics in high-risk groups. However, studies on inequalities stratified by the Socio-demographic Index (SDI) and future trend predictions for these groups are still lacking.MethodsBased on the GBD 2021 database, we extracted data on HZ incidence and disability-adjusted life years (DALYs) among individuals aged ≥ 50 years in 204 countries and regions from 1990 to 2021. We used the Age-Period-Cohort (APC) model, inequality indices (SII/CI), frontier analysis, and BPAC prediction models, combined with stratification by the Socio-demographic Index (SDI) and geography, to explore the spatiotemporal dynamics and driving factors of the disease burden and predict trends up to 2035.ResultsFrom 1990 to 2021, the global incidence of HZ in high-risk groups increased from 5.48 million to 12.49 million cases (a 128% increase), with the age-standardized incidence rate (ASIR) rising slightly from 640.9 to 654.1 per 100,000 (EAPC 0.1%). DALYs increased from 141,000 to 220,000 (a 56.8% increase), while the age-standardized DALY rate (ASDR) decreased from 18.27 to 11.90 per 100,000 (EAPC − 1.45%). SDI stratification revealed a pattern of “higher incidence but lower burden in high SDI regions, and lower incidence but higher burden in low SDI regions.” Women and individuals aged ≥ 90 years had the highest burden. APC analysis showed that age was the core driving factor, with the period effect indicating a peak in incidence around 2005 and a continuous decline in DALYs due to medical interventions. The cohort effect showed a lower disease burden in younger cohorts. The number of cases is predicted to continue rising by 2035, while the ASIR will fluctuate at a high level.ConclusionThe disease burden of HZ in high-risk groups is driven by multiple factors, including aging, gender, SDI, and access to medical care, showing significant multidimensional disparities. Stratified precision prevention and control measures (strengthening surveillance and elderly interventions in high SDI regions, and supplementing basic medical resources in low–middle SDI regions), widespread vaccination, and international resource redistribution are needed to curb the rising disease burden and promote global health equity.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12985-025-03044-7. Background Herpes zoster(HZ) is triggered by the reactivation of the varicella - zoster virus (VZV) [1]. As an infectious disease attracting global attention, HZ not only causes skin lesions but is also frequently accompanied by severe complications such as postherpetic neuralgia [2]. Related research indicates that immune aging elevates the risk of viral reactivation and the risk of HZ surges significantly after 50 years old and the severity of the disease tends to increase with advancing age [3]. Thus, individuals aged 50 years and older are high - risk groups for HZ infection.Over 95% of adults aged 50 years or older have been infected with VZV, and the estimated lifetime risk of developing HZ is approximately 30% [4]. Postherpetic neuralgia (PHN), the most common complication of HZ, affects 10% to 18% of patients and is characterized by chronic neuropathic pain that significantly impairs quality of life [5]. The age-related decline in VZV-specific T-cell immunity leads to an exponential increase in HZ incidence, rising from 4.0 to 4.5 cases per 1,000 person-years in the general population to more than 11.0 cases per 1,000 person-years among individuals aged 80 years or older [6, 7]. Empirical data indicates that the annual medical expenditure attributable to herpes zoster in Germany, France, and the United Kingdom is as high as 199.87 million euros, 205.7 million euros, and 271.21 million euros, respectively [8].In the United States, the medical costs associated with herpes zoster and its complications exceed 2.4 billion US dollars [9]. The Global Burden of Disease (GBD) study offers multi - dimensional data support across different regions and extended time series for a systematic analysis of the epidemic characteristics of diseases. By examining core metrics such as incidence and disability - adjusted life years (DALYs), it is possible to accurately characterize the distribution pattern and evolution laws of the disease burden of herpes zoster among the high - risk groups (individuals aged 50 years and older). Previous studies have uncovered the age, gender, and geographical heterogeneities in the disease burden of HZ [10]. Nevertheless, in the context of persistent global health inequalities and rapid population aging, comprehensive analyses regarding the high - risk groups on the inequalities stratified by the sociodemographic index (SDI), age-period-cohort (APC) model effects and future trend predictions are lacking. Thus, the study focuses on the high-risk groups, evaluating their driving factors, evolution trajectory and key points of prevention and control of HZ disease burden, which will be beneficial for optimizing the allocation of medical resources and formulating precise prevention and control strategies. Based on the GBD 2021 database, this study systematically dissects the characteristics of age, gender, region, and SDI stratification regarding the incidence and DALYs of HZ among high-risk groups for the first time. By using the APC model, inequality analysis and predictive models, this research delves deep into the evolution laws and underlying driving factors of the disease burden within high-risk groups. Moreover, this study reveals the multi-dimensional heterogeneities and future trends of the disease burden of HZ among high-risk groups, which not only furnishes a scientific foundation for formulating stratified prevention and control strategies but also contributes to the establishment of a more targeted health management system for HZ among high-risk groups. Methods Data acquisition and sources This study utilized data from the GBD 2021 database, which provides comprehensive estimates of the burden of 371 diseases and injuries and 88 risk factors in 204 countries and regions from 1990 to 2021 [11]. Data specific to HZ in high-risk groups were extracted, including incidence, prevalence, and DALYs. Years of life lost (YLLs) are not applicable to HZ in high-risk groups because it is a non - fatal disease. Due to the lack of YLLs, the DALYsfor HZ amonghigh-risk groups are equivalent to years lived with disability (YLDs). This dataset can be accessed through the Global Health Data Exchange (GHDx) platform (https://ghdx.healthdata.org/gbd − 2021/sources). SDI data and 21 regions were also included to assess the impact of socioeconomic factors on the burden of AA. All disease burden estimates were presented in the form of counts, age - standardized rates, and age - standardized percentage changes from 1990 to 2021. These results were accompanied by 95% uncertainty intervals (UI) to account for variability and ensure the reliability of the results. Global and regional disease burden analysis Using R (version 4.4.1) with the ggplot2 and sf packages we constructed a comparative analysis between global and regional levels based on geographic groupings (covering 204 countries and 21 regions) from the GBD database. The data were were visualized to assess global trends in HZ among high-risk groups. Population analysis and age‑standardized rate (ASR) We extracted data for individuals aged ≥ 50 years, categorized into ten age groups: 50–54, 55–59, 60–64, 65–69, 70–74, 75–79, 80–84, 85–89, 90–94 and ≥ 95 years. Statistical significance of trends was determined using 95% confidence intervals (CI). Moreover, ASR were calculated based on the world standard population reported in the Global Burden of Disease Study 2021 to enable cross-regional comparisons. The reference population data we used was from the GBD 2021 which provides age-specific population weights, as documented in the GBD 2021 Demographics 1950–2021 dataset. AAPC (Average Annual Percentage Change) represents the annualized change percentage, derived by converting the weighted average of slope coefficients from the underlying joinpoint regression model spanning 1990 to 2021. In this study, a rate was considered to exhibit an upward trend if both the estimated annual percentage change and its 95% CI exceeded 0, or a downward trend if both values were below 0. Time trend analysis We employed the segmented package in R languageand piecewise linear regression model to analyze temporal trends in the incidence and DALYs of HZ from 1990 to 2021. Results are presented as APC and AAPC with 95% CI to determine statistical significance. The AAPC was calculated to quantify the average rate of change over time. Sociodemographic index (SDI) analysis To assess the relationship between the SDI and disease burden, we calculated SDI-specific rates. Disease burden comparisons across socioeconomic strata were performed using five SDI levels: low, low-middle, middle, middle-high, and high. Data manipulation and visualization were implemented using the dplyr and ggplot2 R packages. Frontier analysis We employed frontier analysis to quantify HZ burden associated with sociodemographic development and identify the theoretically achievable minimum ASR based on SDI. The efficiency gap metric was used to measure disparities between observed health outcomes and theoretical optima (frontier points) under given SDI conditions. A larger efficiency gap indicates lower health system efficiency. The frontier DALYs rate represents the minimum achievable value for each country or region according to its SDI level. The age-period-cohort (APC) model The analysis utilised a segmented linear regression model, the inflection point of which was determined by minimising the weighted mean square error. This model was used to identify significant temporal breakpoints and to calculate the Average Annual Percentage Change and Annual Percentage Change (APC) within the resulting segments [12]. APC model is a statistical framework designed to disentangle the complex interplay of three temporal dimensions influencing population-level outcomes: age effects (biological and social changes associated with aging), period effects (external events contemporaneously affecting all age groups, such as policy changes, economic recessions, or pandemics), and cohort effects (the unique experiences shared by individuals born in the same time interval, shaping their lifelong health or social trajectories, which remains a powerful tool in epidemiology, demography, sociology, and public health for understanding the underlying temporal drivers of trends in outcomes like disease incidence, mortality, social behaviors, or economic indicators. The Bayesian age–period–cohort (BAPC) model, implemented via integrated nested Laplace approximations, projected future trends by disentangling age, period, and cohort effects [13]. Slope index of inequality (SII) and the concentration index (CI) analysis In this study, SII and CI, as defined by the World Health Organization (WHO), were employed to assess inequalities in DALYs, using the SDI as the basis for ranking. The SII was calculated by performing a weighted linear regression of DALYs rates against the midpoint of the cumulative population distribution ranked by SDI. Each country or territory was weighted according to its population size to reflect its proportional contribution to the total population. Data from 204 countries and territories were compared to evaluate changes in health inequalities between 1990 and 2021. Furthermore, the CI, measuring relative inequality, was derived by plotting the cumulative distribution of DALYs against the cumulative distribution of SDI, with weighting applied to reflect the population size of each country or territory. The CI value was obtained by matching the cumulative proportion of DALYs to the cumulative population distribution ranked by SDI and numerically integrating the area under the Lorenz concentration curve. Statistical analysis All statistical analyses and visualizations were performed using R (version 4.4.1). Descriptive statistics are reported as means with their 95% uncertainty intervals (UI) or 95% CI.For trend analyses, p-values below 0.05 were considered statistically significant. Results HZ global disease burden trends The age-standardized incidence rate (ASIR) of HZ among high-risk groups in 2021 was shown in the global map (Fig. 1A). From 1990 to 2021, the total global incidence of HZ rose from approximately 5,482,944.69 cases to around 12,492,173.71 cases, registering a growth rate of 128%. This indicates that a substantial number of individuals remain at risk of this disease (Table 1). The ASIR increased slightly from 640.9 per 100,000 in 1990 to 654.1 per 100,000, with an estimated annual percentage change EAPC of 0.1% (95%CI: 0.16 to 0.14), indicating a slow but statistically significant increase in incidence. Fig. 1Overview of global burden of HZ among high-risk groups distribution. A Heatmap of global burden of incidence; B Heatmap of global burden of DALYs Overview of global burden of HZ among high-risk groups distribution. A Heatmap of global burden of incidence; B Heatmap of global burden of DALYs Meanwhile, the number of HZ DALYs cases among high-risk groups increased from 140,647.09 (95% CI: 111,522.01, 178,471.27) in 1990 to 220,468.16 (95% CI: 156,900.91, 300,327.51) in 2021, representing a 56.8% increase, which reflects a substantial augmentation in the absolute burden. The age - standardized disability - adjusted life years rate (ASDR) declined from 18.27 cases per 100,000 population in 1990 to 11.90 cases per 100,000 population in 2021 (Fig. 1B), with an estimated annual percentage change EAPCs of −1.45(95%CI: −1.50,−1.39) (Table 2). Table 1.Numbers,ASR and EAPC ASR for incidence of Herpes zoster in high-risk groups (1990-2021) from Global Burden of Disease Table 1.Numbers,ASR and EAPC ASR for incidence of Herpes zoster in high-risk groups (1990-2021) from Global Burden of Disease Numbers,ASR and EAPC ASR for incidence of Herpes zoster in high-risk groups (1990-2021) from Global Burden of Disease Regional trends by sociodemographic index (SDI) In 2021, in regions with a medium SDI, the incidence among high-risk groups was the highest, amounting to 4,104,888.93 cases (95% CI: 2,761,869.49, 5,718,974.02). In contrast, the number of incident cases among the high-risk groups was the lowest in low SDI regions, reaching 617,744.71 cases (95% CI: 411,946.89, 854,791.99). In the meantime, the age-standardized incidence rate (ASIR) was the highest in the high SDI regions at 712.32(95%CI: 480.42,995.62), and the lowest in the low SDI regions at 579.20(95%CI: 385.14,807.95). The estimated average annual percentage change (EAPC) in the age - standardized incidence rate was the highest in regions with a high SDI, at 0.29 (95% CI: 0.17, 0.41), and the lowest in regions with a medium SDI, at − 0.07 (95% CI: − 0.16, − 0.02) (Table 1). However, the ASDR was the highest in the low SDI regions, at 34.46(95%CI: 26.61,43.96), and the lowest in the high SDI regions, at 7.31(95%CI: 4.30,11.24) in 2021.While the average annual percentage change (EAPC) in ASDR was negative across all SDI regions. The lowest EAPCs was observed in medium SDI regions, at − 2.56 (95% [CI]: − 2.65, − 2.47) and the EAPCs in low SDI and high SDI regions were − 1.70 (95% CI: − 1.77, − 1.63) and − 0.17 (95% CI: − 0.24, − 0.11), respectively. The above studies suggested that SDI stratification drives disease trend differentiation, with a global pattern of “increased incidence but relieved burden in high SDI regions and moderate incidence but accumulated burden in low SDI regions”. Table 2.Numbers,ASR and EAPC ASR for DALYs of Herpes zoster in high-risk groups (1990-2021) from Global Burden of Disease Table 2.Numbers,ASR and EAPC ASR for DALYs of Herpes zoster in high-risk groups (1990-2021) from Global Burden of Disease Numbers,ASR and EAPC ASR for DALYs of Herpes zoster in high-risk groups (1990-2021) from Global Burden of Disease Trends of HZ Diease burden in different geographical regions Based on the GBD database, this study conducted an in-depth exploration of the spatio-temporal evolution trends and regional clustering characteristics of the incidence and DALYs of HZ among high-risk groups (Fig. 2). The EAPCs of the HZ incidence and DALYs among high-risk groups in different regions were presented in the form of a bar chart, as shown is Fig. 2A.In terms of incidence rate, the EAPC is significant heterogeneity across different regions globally. In developed regions such as High-income North America and High-income Asia Pacific, the EAPC of incidence rate is positive and relatively high, at 0.39 (95% CI: 0.09, 0.68) and 0.60 (95% CI: 0.44, 0.75) respectively. This indicates that the incidence of HZ among high-risk groups is on the rise year by year in these regions, which may be related to the aging population and the improved disease surveillance system that captures more cases; while in some parts of sub-Saharan Africa (such as Eastern Sub-Saharan Africa), the EAPC of incidence is close to 0 or negative, reflecting that the growth trend of the disease is relatively stable or due to limited medical resources and a high rate of missed diagnosis, the true incidence trend has not been fully revealed. In the DALYs dimension, high SDI regions (such as high-income North America and Western Europe) have mostly negative EAPCs, indicating that with the optimization of medical interventions (such as antiviral treatment and pain management), disease-induced loss of healthy life expectancy has decreased year by year. In conclusion, high SDI regions need to enhance immune protection for the elderly and optimize disease surveillance and low SDI regions should focus on supplementing medical resources and expanding intervention coverage. Fig. 2Overview of global burden of HZ among high-risk groups EAPC change. A Bar plot of EAPC of incidence and DALYs; B Cluster analysis of EAPC of incidence and DALYs. EAPCestimated annual percentage change Overview of global burden of HZ among high-risk groups EAPC change. A Bar plot of EAPC of incidence and DALYs; B Cluster analysis of EAPC of incidence and DALYs. EAPCestimated annual percentage change Clinical characteristics of HZ patients among high-risk groups Based on the age and gender composition of the patients, we also analyzed the trends of incidence and DALYs among high-risk groups. The stratified comparison of incidence and DALYs showed the incidence and disability rate of HZ among high-risk groups gradually increase with age all over the world. The age group of ≥ 90 years old takes up a prominent share in both the incidence proportion and disease burden proportion. For instance, in high-income North America, the incidence proportion of the ≥ 90 years old group is 14.3% and the DALYs proportion is 44.5%, which clearly reflects the strong correlation between age growth and the incidence of HZ and health loss among high-risk groups (Fig. 3A-B). Figure 3C-Dfurther explores from the gender dimension, which suggested that in most regions around the world, the HZ incidence and DALYs among women are both higher than those among men. The number of cases and disabilities of HZ among high-risk groups in the 100,000 population by age and gender shows that with increasing age (Fig. 3E-F), in summary, the burden of herpes zoster disease presents the distribution characteristics of “heavier burden with age, dominant female population, significant regional differences” This provides a key basis for precisely formulating stratified prevention and control strategies and optimizing the allocation of medical resources. Identifying the elderly and female groups as key intervention targets helps to effectively reduce the impact of the disease on population health and provides strong data support for public health decision-making. Fig. 3Impact of sociodemographic transition on the burden of HZ among high-risk groups. A Composition ratio of incidence and DALYs rates in different age groups(individuals ≥ 50 years) in 2021, globally and by 21 regions; B Barplot of proportion of disease burden in male and female groups in the 27 regions defined by GBD 2021; C Number of incidence and DALYs cases globally, incidence and DALYs rate of HZ among high-risk groups per 100 000 population, by age and sex in 2021. Lines indicate prevalent case with 95% UIs for men and women Impact of sociodemographic transition on the burden of HZ among high-risk groups. A Composition ratio of incidence and DALYs rates in different age groups(individuals ≥ 50 years) in 2021, globally and by 21 regions; B Barplot of proportion of disease burden in male and female groups in the 27 regions defined by GBD 2021; C Number of incidence and DALYs cases globally, incidence and DALYs rate of HZ among high-risk groups per 100 000 population, by age and sex in 2021. Lines indicate prevalent case with 95% UIs for men and women Spatio-temporal trends of HZ burden among high-risk groups from 1990 to 2020 Globally, the incidence rate rose rapidly from about 640 per 100,000 during 2000–2005 to a peak of about 670 per 100,000, then briefly declined and gradually rebounded after 2010, remaining fluctuating at around 650 per 100,000 in 2021; The corresponding ASDR showed a continuous decline from 1900 to 2021, reflecting that although there were short-term fluctuations in incidence rates, with the improvement of medical standards, the long-term reduction in healthy life expectancy caused by HZ among high-risk groups (Fig. 4A-B). Over the past 30 years, although the overall trend of incidence among different age groups has been relatively stable, the older the age, the higher the incidence, suggesting that the older population faces a higher risk of disease over the long term. Meanwhile, with age, ASDR in the ≥ 75 age group shows a slow decline but remains at a high level, highlighting the ongoing impact of shingles on the healthy lifespan of the elderly (Fig. 4C-D). During 1990–2020, the incidence baseline was relatively stable in areas with medium to low SDI, significantly lower than in other areas. In contrast, the incidence rates in the medium and high SDI regions remained high and showed significant fluctuations, with peaks around 2005 and 2015. The sudden increase in incidence rates may be related to the improvement in disease detection capacity in the region; Over time, ASDR in all SDI regions showed a downward trend, but in the medium-low SDI region, it was at a higher level, significantly higher than the global level, while in the high SDI region, it was at the lowest level (Fig. 4E-F). Fig. 4Temporal trends of incidence and DALYs for HZamong high-risk groups. A AAPC for incidence rates from 1990 to 2021. B Trends of incidence and DALYs rates by age groups from 1990 to 2021. C Tends of ASIR and DALYs from 1990 to 2021. SDI sociodemographic index Temporal trends of incidence and DALYs for HZamong high-risk groups. A AAPC for incidence rates from 1990 to 2021. B Trends of incidence and DALYs rates by age groups from 1990 to 2021. C Tends of ASIR and DALYs from 1990 to 2021. SDI sociodemographic index The Temporal trends and driving factors of global incidence and dalys for HZ among high-risk groups To overcome the inherent potential bias of non-parametric methods such as ASR and EAPC, we further explored the longitudinal epidemiological trends of HZ among high-risk groups using APC and AAPC methods. The age-specific annual percentage change in HZ incidence (local drift) versus the overall annual percentage change (net drift) fluctuated with age. The net drift was 0.095, with the highest risk observed at 67.5 years (drift value 0.115) and 92.5 years (drift value 0.13) (Fig. 5A); Based on the results of the age effect, the incidence and disability rate of HZ among high-risk groups showed a gradually increasing trend (Fig. 5B, F), suggesting that age is a core driver of the onset and disease burden of HZ; In terms of the period effect results, the global incidence risk of HZ shows fluctuating characteristics over time, reaching a peak of 1.0 around 2005 and then gradually decreasing, while the period effect of DALYs shows a downward trend, reflecting that with the advancement of medical technology (the popularization of antiviral treatment, the optimization of pain management, etc.), The loss of healthy life expectancy caused by the disease decreased gradually over the same period (Fig. 5C and G). Meanwhile, cohort effect analysis showed that the incidence risk in the birth cohort fluctuated but generally increased, while the DALYs cohort effect was the opposite, with a lower disease burden in the younger cohort, reflecting the long-term benefits of medical advancements for the health of offspring (Fig. 5D and H). Fig. 5Age–period–cohort (APC) analysis and decomposition analysis to estimate temporal trends and driving factors of global incidence and DALYs for HZ in among high-risk groups. A–D Net drifts and local drifts of incidence rate by age group A impact of age effects on incidence rate B rate ratio of period C and cohort D effects; E–H Net drifts and local drifts of DALYs rate by age group A impact of age effects on incidence rate B rate ratio of period C and cohort D effects; I–J Population-level determinant changes in aging, population growth, and epidemiological changes for ischemic stroke in young adults globally and in various SDI regions from 1990 to 2021. Black dots represent the total change contributed by all three components. A positive value for each component indicates a corresponding positive contribution in incidence and DALYs, and a negative value indicates a corresponding negative contribution in incidence and DALYs. SDI sociodemographic index Age–period–cohort (APC) analysis and decomposition analysis to estimate temporal trends and driving factors of global incidence and DALYs for HZ in among high-risk groups. A–D Net drifts and local drifts of incidence rate by age group A impact of age effects on incidence rate B rate ratio of period C and cohort D effects; E–H Net drifts and local drifts of DALYs rate by age group A impact of age effects on incidence rate B rate ratio of period C and cohort D effects; I–J Population-level determinant changes in aging, population growth, and epidemiological changes for ischemic stroke in young adults globally and in various SDI regions from 1990 to 2021. Black dots represent the total change contributed by all three components. A positive value for each component indicates a corresponding positive contribution in incidence and DALYs, and a negative value indicates a corresponding negative contribution in incidence and DALYs. SDI sociodemographic index From a macro perspective, considering population growth, population aging, and epidemiological changes helps us understand the main drivers of changes in disease incidence and disability rates in different regions and globally. There will also be differences among regions with different SDI. In medium SDI regions, the increase in incidence is mainly driven by population growth, while in medium to high SDI regions and high SDI regions, the impact of population aging is gradually increasing, and together with population growth, it contributes to the disease burden of HZ among high-risk groups. Aging remains the main contributor to disability rates in High SDI areas such as high-income Asia Pacific; In Low-middle SDI regions (such as low-middle SDI), the combined effect of demographic factors and epidemiological changes (such as disease spectrum adjustment, burden reconfiguration after increased medical accessibility), and the impact of epidemiological changes in some areas (blue bars) on DALYs is gradually emerging, suggesting that under SDI stratification Disease burden drivers evolve with the stage of socio-economic development. High SDI areas focus on health management for the elderly, while low - and medium-SDI areas need to balance the dual effects of population growth and medical improvement (Fig. 5I and J). Prediction analysis using AAPC and age–period–cohort analysis ASIR in 204 countries and regions by SDI increased as SDI increased. Notably, the GBD framework to conduct an in-depth analysis of ASIR curve did not continue to rise with the increase in SDI, but rather reached a plateau and declined slightly in regions with medium to high SDI. Followed by rapid growth in high SDI regions (Fig. 6A, B); Meanwhile, this study used methods from the field of health inequalities, focusing on the differences in incidence rates among different populations (distinguished by sociodemographic index, geographical region, age and gender differences). To quantify inequality more precisely, we used the Slope Index of Inequality (SII) and Concentration Index of Inequality (CII) indicators to quantify differences in the burden of disease at socioeconomic, geographical, and demographic levels Revealing key drivers of health inequality, with incidence rates in some countries significantly above or below the regression line, which may suggest significant health inequality in those countries (Fig. 6C, D). Further, countries with higher incidence rates than expected could be identified through a frontier analysis based on SDIand an analysis of inequality associated with SDI. As shown in Fig. 6, the top 15 countries with the greatest effective differences are: Republic of Korea, Israel, Brunei Darussalam, Singapore, Taiwan(Province of China), France, Cook Islands Malaysia, Guam, Northern Mariana Islands, Japan, Palau, Germany, Spain, Belgium imply that these countries are at a disadvantage from the front and may indicate that they have specific health problems in some respects. Some countries scored very close to the baseline, such as Somalia, Chad Mali, Niger, South Sudan (Fig. 6E, F). Fig. 6SDI-related inequalities and frontier analysis of HZ among high-risk groups in 204 countries and territories. A, B ASIR and DALYs rates of 204 countries and territories by SDI in 2021; C, D Health inequality regression curvesand concentration curves for the incidence rates of ischemic stroke in young adults across the world, from 1990 to 2021; E The color change from light blue (1990) to dark blue (2021) represents the change in years. The frontier is delineated in solid black color; countries and territories are represented as dots. F The top 15 countries and territories with the largest effective difference (largest ASIR of herpes zoster in individulas ≥ 50 years gap from the frontier) are labeled in black; examples of frontier countries with low SDI(< 0.5) and low effective difference are labeled in blue, and examples of countries and territories with high SDI (> 0.85) and relatively high effective difference for their level of development are labeled in red. Red dots indicate an increase in ASIR of H in HZ among high-risk groups from 1990 to 2021; blue dots indicate a decrease in ASIR of HZ among high-risk groups from 1990 to 2021. The direction of ASR change from 1990 to 2021 is indicated by the color of the dots, with orange dots representing decreases and blue dots representing increases. SDI sociodemographic Index SDI-related inequalities and frontier analysis of HZ among high-risk groups in 204 countries and territories. A, B ASIR and DALYs rates of 204 countries and territories by SDI in 2021; C, D Health inequality regression curvesand concentration curves for the incidence rates of ischemic stroke in young adults across the world, from 1990 to 2021; E The color change from light blue (1990) to dark blue (2021) represents the change in years. The frontier is delineated in solid black color; countries and territories are represented as dots. F The top 15 countries and territories with the largest effective difference (largest ASIR of herpes zoster in individulas ≥ 50 years gap from the frontier) are labeled in black; examples of frontier countries with low SDI(< 0.5) and low effective difference are labeled in blue, and examples of countries and territories with high SDI (> 0.85) and relatively high effective difference for their level of development are labeled in red. Red dots indicate an increase in ASIR of H in HZ among high-risk groups from 1990 to 2021; blue dots indicate a decrease in ASIR of HZ among high-risk groups from 1990 to 2021. The direction of ASR change from 1990 to 2021 is indicated by the color of the dots, with orange dots representing decreases and blue dots representing increases. SDI sociodemographic Index We also used the BPAC model to predict the incidence of HZ and ASIR among high-risk groups worldwide from 1990 to 2035. As shown in Fig. 7, the X-axis of the chart on the left represents the year and the Y-axis represents the number of new cases in millions. The red dotted line represents the predicted trend of the number of cases, and the gray area represents the range of uncertainty in the prediction. The graph shows that the number of cases has continued to rise over time from 1990 to 2020 and is expected to maintain the current growth rate and reach a higher level by 2035 (Fig. 7A), reflecting that the number of cases of shingles is driven by factors such as population growth and aging, and the future growth trend is significant, suggesting that disease control needs to deal with the pressure of the continuous increase in the number of cases. The X-axis on the right side of the chart represents the period from 1990 to 2035, and the Y-axis represents ASIR per 100,000 people. The red dotted line in the graph represents the predicted trend of ASIR, while the gray dots and areas represent the confidence intervals of historical data and predictions. It can be seen from the graph that ASIR rose to a peak in 2005 and then gradually declined, stabilized in 2010 and then fluctuated upward, reached a peak in 2015 and then slowly declined until 2018 and then gradually stabilized. After 2020, the growth rate of ASIR will remain stable and rise slowly until 2035 (Fig. 7B). Fig. 7Global change trends of HZ among high-risk groups from 1990 to 2021, and its predicted trends between 2022 and 2035. Temporal trend in the ASR (A) and numbers (B) of incidence cases from 1990 to 2035 for HZ among high-risk groups. Black dots represent observed ASR, and red dots represent ASR predicted by the BAPC model; purple-shaded areas represent predicted intervals, expressed as confidence intervals with confidence level of 95% (the outermost), 80%, 70%, 60%, and 50% (the innermost). ASR age-standardized rates Global change trends of HZ among high-risk groups from 1990 to 2021, and its predicted trends between 2022 and 2035. Temporal trend in the ASR (A) and numbers (B) of incidence cases from 1990 to 2035 for HZ among high-risk groups. Black dots represent observed ASR, and red dots represent ASR predicted by the BAPC model; purple-shaded areas represent predicted intervals, expressed as confidence intervals with confidence level of 95% (the outermost), 80%, 70%, 60%, and 50% (the innermost). ASR age-standardized rates Discussion The present study leverages the GBD 2021 data to meticulously examine the incidence and DALYs attributable to HZ among high-risk populations (individuals aged ≥ 50 years) across a 31-year span from 1990 to 2021. Our analysis underscores that age, gender disparities, and regional variations in the SDI are pivotal determinants of the HZ disease burden. Over this period, there was a substantial 128% surge in the global incidence of HZ, escalating from approximately 5.48 million cases to around 12.49 million cases. While the a ASIR exhibited a modest upward trajectory, the ASDR concurrently demonstrated a notable decline. These divergent trends highlight the intricate interplay between demographic shifts, socioeconomic development, and healthcare accessibility in shaping the epidemiological landscape of HZ. Our age-effect analysis revealed a pronounced increase in HZ incidence and DALYs with advancing age, firmly establishing the elderly as the core high-risk cohort [14, 15]. Previous studies have confirmed that advanced age is one of the primary risk factors for HZ, which may be attributed to the fact that the immune system in the elderly is more fragile compared to that of younger individuals [16]. The human body primarily relies on the cellular immune system to combat the varicella-zoster virus (VZV). Compared to younger people, the elderly have a reduced number of VZV-specific T cells, slower responses, and a higher proportion of functionally abnormal T cells, all of which are more likely to trigger viral reactivation [17, 18]. With the continuous rise in the proportion of the elderly population, significant challenges will persist over the next few decades [19]. Therefore, it is imperative to develop targeted protective strategies for the elderly. Additionally, the higher incidence and DALYs observed in women compared to men corroborate the complex role of estrogen in immunomodulation, aligning with prior research that identifies female gender as an independent risk factor for HZ.This is consistent with previous research results, indicating that female gender is an independent risk factor for herpes zoster [20]. A study on the incidence of HZ among Chinese people showed that the cumulative incidence and average annual incidence rates were higher in women than in men, with the highest cumulative incidence rate among those aged 80 and above [21, 22]. SDI analysis revealed heterogeneity in the association between socioeconomic development and the onset and burden of shingles. Higher incidence in high SDI regions is directly associated with improved disease surveillance (reduced missed diagnoses) and accelerated aging (increased proportion of the elderly population); The burden of DALYs decreases with the increase in SDI, highlighting the mitigation of health depletion by the accessibility of medical resources such as antiviral drugs and pain management systems [23]. Low - and medium-SDI regions face the dilemma of “inadequate disease surveillance and lagging burden relief”, which is consistent with the conclusion of the latest study that “global health inequality persists and the shortage of medical resources in low SDI regions restricts disease prevention and control” [24]. This requires a stratified strategy: high SDI regions focus on the “monitoring-intervention” closed loop (such as using big data to warn of the peak incidence), while medium - and low SDI regions prioritize supplementing basic medical resources (vaccination networks, primary care capabilities) to narrow the gap of unequal disease burden. The prediction results of the continuous increase in the number of cases and the slow rise in age-standardized morbidity reflect the game between population aging and medical intervention. On the one hand, the proportion of the global elderly population is projected to reach 22% by 2050, and aging drives an increase in the number of cases; On the other hand, medical advancements (such as widespread vaccination and early diagnosis and treatment) have slowed the rate of increase in morbidity [25]. HZ vaccine may be the key to solving this dilemma. At present, there are mainly two types of herpes zoster vaccines: the live attenuated vaccine Zostavax, which was approved by the FDA in 2006, and the recombinant vaccine Shingrix, which was approved in 2017. Multiple studies have confirmed the significant efficacy of the vaccines — clinical trials have shown that the live attenuated vaccine Zostavax can reduce the incidence of herpes zoster by 51.3% and the incidence of postherpetic neuralgia by 66.5% among people aged 60 and above, while also demonstrating good safety [26, 27]. According to the data from Ryan Thompson’s team, since the introduction of the live attenuated vaccine in 2007, the incidence of HZ among people aged 60 and above in the US has been declining year by year. In addition, with the development of new vaccines, their protective efficacy is still improving. The research by Hymal Lal and colleagues has revealed that the recombinant subunit vaccine confers a striking preventive efficacy of 97.2% in individuals aged 50 years and older, thereby substantially diminishing the incidence of herpes zoster [28]. At the same time, the severity of the disease in patients who continue to have the disease after vaccination is also significantly reduced. Studies have found that vaccination against herpes zoster can even reduce the risk of stroke [29]. Unlike the live attenuated vaccine, the recombinant subunit vaccine can also provide protection for immunocompromised individuals with underlying diseases [30]. More and more people have the opportunity to be protected by vaccines. The EAPC cluster shows a pattern of “increased incidence but decreased burden in high SDI areas and stable incidence but increased burden in low SDI areas”, which is deeply bound to regional medical resources and population structure. High-income regions have reduced DALYs through advanced diagnostic and therapeutic techniques such as multimodal analgesia and immunomodulatory therapy, but aging has pushed up the incidence; Low-income regions, with scarce medical resources, have difficulty alleviating the disease burden, which aligns with the call for global health governance to focus on resource redistribution. The next step is to tailor regional plans based on clustering characteristics: in high SDI regions, strengthen the synergy of “disease surveillance - elderly intervention”; in low SDI regions, promote the implementation of “resource allocation - basic prevention and control”, and narrow regional disparities through international medical assistance and technology transfer. Limitations This study relies on secondary data from the GBD database and has limitations such as uneven regional data coverage and inconsistent monitoring quality in some areas, which may affect the universality of the results. In the future, multi-center cohort studies will be needed to supplement data from primary medical care scenarios; At the same time, explore the impact of genetic polymorphisms on disease burden, deepen the “genetice-environment-society” interaction mechanism, and provide molecular-level evidence for precise prevention and control. Conclusion In summary, the current drivers of the HZ disease burden include population aging, gender differences, and imbalances in regional socio-economic development. High-SDI regions need to focus on elderly health management and disease surveillance to address the rising incidence, while low-SDI regions should prioritize supplementing basic medical resources and enhancing disease monitoring and intervention capabilities to alleviate the heavy disease burden. Future projections show that by 2035, the global number of HZ cases will continue to rise, while the ASIRwill fluctuate at a high level. This suggests that future prevention and control efforts need to balance the increasing number of cases driven by population aging with the reduction of incidence rates through medical advancements. Meanwhile, international redistribution of medical resources and technology transfer will be key to narrowing regional disparities in disease burden. Supplementary Information Supplementary Material 1 Supplementary Material 1