The age-standardized rates (ASR), mortality rates, percentages, and DALYs were extracted from the GBD 2019 study. Crude incidence rates (e.g., prevalence, incidence, mortality) are fundamental indicators for tracking disease trends. However, differences in population age structure can cause heterogeneity in the burden of gastrointestinal ulcers across groups. To ensure comparability, we adjusted the crude rates by age structure to obtain age-standardized rates (ASR). The age-standardized prevalence rate (ASPR), age-standardized incidence rate (ASIR), age-standardized DALY rate (ASDR), and age-standardized death rate (ASDER) were then used to estimate the burden of gastrointestinal ulcers in adolescents aged 10–24.

We used a generalized linear model with a Gaussian distribution to calculate the estimated annual percentage change (EAPC) of the prevalence rate, incidence rate, DALY rate, and death rate to quantify global disease trends. If the lower limit of the EAPC and its 95% confidence interval (CI) is > 0, the disease burden is considered to have increased. Conversely, if the upper limit of the EAPC and its 95% CI is < 0, a decreasing trend is observed.

Additionally, the ARIMA (Autoregressive Integrated Moving Average) model, a statistical tool for forecasting and analyzing time series data, was employed in our study. ARIMA plays a crucial role by predicting disease trends, which aids in public health policy-making and resource allocation. The model can also predict missing data based on existing data, improving dataset completeness and enhancing the accuracy of the research. ARIMA helps identify outliers, ensuring data quality and accurate trend interpretation. Thus, it provides in-depth insights into disease trends and enhances analytical precision, making it valuable for global disease burden research.

The Exponential Smoothing (ES) model, another time series forecasting tool, was also used in GBD analysis. The ES model is well-suited for predicting disease trends and evaluating public health intervention.

In 2019, there were 958,842 (95% UI: 639,698–1,371,106) prevalent cases of gastrointestinal ulcers globally, with an age-standardized prevalence rate (ASPR) of 51.5 per 100,000 (95% UI: 34.36–73.64), reflecting a 6.6% decrease since 1990. Additionally, 407,850 (95% UI: 260,513–577,751) new cases were reported, with an age-standardized incidence rate (ASIR) of 21.91 per 100,000 (95% UI: 13.99–31.03), a 3.8% decline since 1990. The number of disability-adjusted life years (DALYs) related to gastrointestinal ulcers was 363,862 (95% UI: 309,793–422,230), with an age-standardized DALY rate (ASDR) of 19.54 (95% UI: 16.64–22.68), showing a significant 50.68% reduction since 1990. There were 4,404 (95% UI: 3,826-5,042) deaths, with an age-standardized death rate (ASDER) of 0.24 (95% UI: 0.21–0.27), a 53.8% decrease from 1990.

In 2019, the highest ASPR, ASIR, ASDR, and ASDER for gastrointestinal ulcers were observed in Kiribati, with 232.47 (95% UI: 172.26–303.4), 69.17 (95% UI: 50.14–93.99), 121.9 (95% UI: 57.99–201.99), and 1.58 (95% UI: 0.65–2.75) per 100,000, respectively. Conversely, the lowest ASPR and ASIR were found in Costa Rica, with 2.75 (95% UI: 1.42–4.71) and 1.24 (95% UI: 0.65–2.13) per 100,000, respectively. Andorra had the lowest ASDR at 0.69 per 100,000 (95% UI: 0.43–1.02), while ASDER was zero in multiple countries.

From 1990 to 2019, the Syrian Arab Republic experienced the largest increases in ASPR (44.8%) and ASIR (40.8%), while Zimbabwe saw the greatest rises in ASDR (39%) and ASDER (42%). In contrast, Bhutan saw a 71% decrease in both ASPR and ASIR, Spain had the largest decrease in ASDR (85.5%), and ASDER dropped to zero in several countries (Supplementary Tables 1–4).

Regions with incidence rates exceeding 30% in 2019 included parts of Africa, Europe, and East Asia, while most of North America had rates between 10 and 20%. Areas with the highest prevalence rates were in Africa, Central Asia, and South Asia, with some regions surpassing 120%, while regions with the lowest prevalence rates, including most of the Americas and Europe, ranged from 0 to 30%. Mortality rates were highest in Africa and lowest in most of North America and Europe, where rates were under 0.2. DALY rates were highest in sub-Saharan Africa and parts of South America, and lowest in most of North America and Europe.

In terms of the number of incident cases, prevalent cases, deaths, and DALYs, populous countries like India and China had relatively high levels. However, India had a globally high number of deaths and DALYs, which warrants attention from local health authorities and the national government. Overall, the disease burden among the 10–24 age group shows significant global disparities, with high levels in some African and Asian countries, and lower levels in many North American, European, and East Asian countries (Figure 1).

Among the 44 GBD regions, the highest age-standardized prevalence rate (ASPR) and incidence rate (ASIR) for gastrointestinal ulcers were in Western Sub-Saharan Africa, with 123.93 cases per 100,000 (95% UI: 85.5–171.52) and 46.7 cases per 100,000 (95% UI: 31.68–65.42), respectively. The highest age-standardized death rate (ASDR) and excess death rate (ASDER) were in Oceania, at 55.86 per 100,000 (95% UI: 39.46–81.64) and 0.72 (95% UI: 0.48–0.98), respectively. Central Latin America had the lowest ASPR and ASIR, at 7.86 (95% UI: 4.95–11.89) and 3.23 (95% UI: 1.99–4.77) per 100,000, respectively. Australasia had the lowest ASDR and ASDER, with 0.81 (95% UI: 0.56–1.17) and 0 (0–0.01), respectively.

The largest increases in ASPR and ASIR were observed in High-income Asia Pacific and Central Europe, with increases of 4.7 and 5.8%, respectively. However, ASDR and ASDER decreased across all regions. The country with the highest reduction in ASPR saw a 36.8% decrease, while Latin America & Caribbean – WB had the highest drop in ASIR at 45%. The greatest decrease in ASDR occurred in East Asia (77%), while the largest decrease in ASDER was in Australasia, with a 100% reduction (Supplementary Tables 1–4).

In 2019, Western Sub-Saharan Africa had the highest incidence and prevalence rates, while Oceania had the highest death and DALY rates. The highest incidence, prevalence, death, and DALY cases were in the World Bank lower middle-income countries and Asia (Figures 2A–H). In regions with limited healthcare systems, notable gender disparities were observed. Females had significantly higher age-standardized prevalence rates (ASPR) of 76.17 per 100,000 (95% UI: 50.85–108.33) compared to males at 53.18 per 100,000 (95% UI: 36.24–75.44). A similar pattern was seen in the age-standardized incidence rates (ASIR), with females at 31.49 per 100,000 (95% UI: 20.85–42.53) and males at 23.72 per 100,000 (95% UI: 15.91–31.83). However, in East Sub-Saharan Africa, males had the highest age-standardized death rates (ASDR) at 33.93 per 100,000 (95% UI: 20.91–48.16) and excess death rates (ASDER) at 0.91 (95% UI: 0.60–1.20), showing a unique gender pattern (Supplementary Figures 1A,B).

In terms of age distribution, while most regions showed an increasing trend with age, West Africa, Oceania, and Sub-Saharan Africa exhibited distinct patterns, with ASPR peaking in the 15–19 age group and ASIR peaking in the 10–14 age group. Notably, in East Sub-Saharan Africa, the 20–24 age group had the highest disease burden, with an ASDER of 1.82 (95% UI: 1.20–2.32) and an ASDR of 127.72 per 100,000 (95% UI: 86.22–163.15) (Supplementary Figures 2A,B).

Among major populous regions, significant gender disparities were seen in the number of incidence cases, prevalence cases, deaths, and DALYs, particularly in high-density population areas.

Among all SDI regions, Low SDI countries had the highest rates of ASPR, ASIR, ASDR, and ASDER for gastrointestinal ulcers, with 82.81 (95% UI: 57.73–115.49), 34.01 (95% UI: 22.73–47.77), 37.71 (95% UI: 31.16–44.64), and 0.47 (95% UI: 0.38–0.56) per 100,000, respectively. In contrast, High SDI countries had the lowest rates, with 27.25 (95% UI: 18.28–38.69), 11.8 (95% UI: 7.62–16.6), 2.38 (95% UI: 1.75–3.36), and 0.01 (95% UI: 0.01–0.02) per 100,000. Across all SDI regions, ASPR, ASIR, ASDR, and ASDER showed a decreasing trend, with the largest declines in ASPR, ASIR, and ASDR in Low-middle SDI countries (26.6, 21.5, and 65.6%, respectively). ASDER decreased the most in High SDI countries, by 83.3%. The reductions in ASPR and ASIR were more pronounced in Low SDI than in Low-middle SDI regions, while ASDR declined faster in Low-middle SDI countries, highlighting a lag in treatment capabilities in Low SDI regions (Supplementary Tables 1–4).

In 2019, the incidence rate, prevalence rate, mortality rate, and DALY rate all increased as economic levels decreased. The number of incidence cases, prevalence cases, deaths, and DALYs was generally higher in Low-middle SDI countries compared to Low SDI regions, with these regions leading in deaths and DALYs (Supplement Figures 3A–H). Integrating the socioeconomic context, the gray lines represent the expected rates for ASPR, ASIR, ASDR, and ASDER based on the 2019 SDI values. The analysis shows that disease burden indicators generally improve as SDI levels increase. However, Kiribati and Vanuatu have significantly higher disease burden indicators than expected, suggesting substantial room for improvement in these regions (Supplementary Figures 4A–D).

In 2019, gender-specific data for digestive ulcers revealed that females had higher age-standardized prevalence rates (ASPR), incidence rates (ASIR), and disability-adjusted life years (ASDR), with values of 60.27 (95% UI: 40.48–85.08), 24.5 (95% UI: 16–34.83), and 20.26 (95% UI: 16.78–24.33), respectively. However, the age-standardized death rate (ASDER) was higher in males, at 0.24 (95% UI: 0.21–0.29). Both genders showed declines in these indicators, but the reductions in ASPR, ASIR, and ASDR were more pronounced in males, at 10.5, 9.3, and 51.1%, respectively (Supplementary Tables 1–4). In contrast, females experienced a larger decrease in ASDER, at 55.7%. Regarding gender differences in digestive ulcers, incidence rates, number of incident cases, prevalence rates, and number of prevalent cases were higher in females than in males. While there was no significant difference in mortality rates and death counts between genders, males had higher rates. The DALY rate and DALY count were also slightly higher in females, but the difference was not significant (Supplementary Figures 5A–H).

Globally, the burden of digestive ulcers among individuals aged 10 to 24 has declined across 204 countries. The EAPC values for prevalence, incidence, DALYs, and mortality rates decreased by −0.59 (95% CI: −0.73 to −0.46), −0.41 (95% CI: −0.51 to −0.31), −2.81 (95% CI: −2.96 to −2.66), and −3.1 (95% UI: −3.28 to −2.92), respectively. Among the 44 GBD regions, Central Europe saw the highest increases in EAPC values for prevalence (0.29; 95% CI: 0.18–0.41) and incidence (0.33; 95% CI: 0.24–0.42). Sub-Saharan Africa had the largest increase in DALYs (0.33; 95% CI: −0.17-0.83), while Southern Sub-Saharan Africa experienced the highest rise in ASDER (0.37; 95% UI: −0.19-0.94). In contrast, Tropical Latin America showed the largest decreases in prevalence (−3.39; 95% CI: −3.69 to −3.09) and incidence (−3.63; 95% CI: −3.97 to −3.29), and East Asia had the greatest decline in DALYs (−5.29; 95% CI: −5.44 to −5.14). Australasia exhibited the largest decrease in ASDER (−7.16; 95% UI: −7.72 to −6.59). In SDI regions, both prevalence and incidence rates decreased, with the most significant reductions in low SDI regions (−1.54; 95% CI: −1.76 to −1.31) and low-middle SDI regions (−1.25; 95% CI: −1.42 to −1.07). Low-middle SDI countries saw the largest decrease in DALYs (−4.18; 95% CI: −4.47 to −3.9), while high SDI regions had the greatest reduction in mortality (−5.42; 95% UI: −5.65 to −5.19). By gender, EAPC values for prevalence, incidence, DALYs, and mortality showed a downward trend, with females experiencing more significant declines in prevalence (−0.64; 95% CI: −0.82 to −0.46), incidence (−0.37; 95% CI: −0.51 to −0.22), DALYs (−2.89; 95% UI: −3.1 to −2.68), and mortality (−3.3; 95% UI: −3.56 to −3.04). Males had a more pronounced decrease in incidence (−0.45; 95% CI: −0.52 to −0.38) (Supplementary Tables 1–4). The EAPC values for prevalence and incidence were negative in most countries, particularly in parts of the Americas and Northern Africa, suggesting a significant decline in incidence. While mortality rates among adolescents remained stable in most regions, some areas in the Americas showed a positive EAPC value, indicating an increase in adolescent mortality. In contrast, negative EAPC values in parts of the Americas and Asia indicated a reduction in health loss due to the disease during the study period. However, some regions saw positive EAPC values, reflecting an increase in health loss (Figure 3).

The overall time trends from 1990 to 2019 describe changes in disease incidence and prevalence, showing a generally stable trend with slight fluctuations but no significant increase or decrease. Similarly, the number of prevalent cases has shown a relatively stable trend, akin to the number of incident cases. Regarding mortality numbers and rates, we can observe a consistent decline year by year, reflecting improvements in mortality rates. The trends for DALYs and DALYs rates also show a yearly decreasing trend (Figure 4). Based on the time descriptions of SDI regions, we can see that high SDI regions generally outperform low SDI regions in all aspects. The incidence and prevalence rates in high SDI regions show a decreasing trend, while in low SDI regions, they are relatively stable or slightly increasing. Mortality rates and DALYs in all SDI classifications are on a declining trend, but the decline is more significant in low SDI regions (Supplementary Figure 6). These trend lines illustrate the changes in incidence (A), prevalence (B), mortality rates (C), and DALYs rates (D) over time by gender. It is noteworthy that females generally have higher values in all indicators, particularly in incidence and prevalence rates. The differences in mortality and DALYs rates between males and females are not as pronounced as in incidence and prevalence rates, but males typically show higher values (Supplementary Figure 7).

The boundary represents the countries or regions with the lowest disease burden, based on their SDI. The effective distance is the gap between a country’s observed burden and the achievable burden according to its SDI. A significant effective difference suggests unrealized opportunities for improvement, such as reducing the incidence of digestive diseases. Using 2019 data on prevalence, incidence, DALYs, and mortality alongside SDI, we estimated the effective difference for each country or region from the boundary. The largest effective differences were found in Kiribati, Vanuatu, Burkina Faso, and the Central African Republic, which, due to their lower development levels, have relatively poor economic conditions. In contrast, countries with the smallest effective differences, such as Switzerland, the United States, and Finland, are in developed regions. This highlights that countries with lower economic levels experience a higher burden of digestive ulcers. In the figure, the black solid line represents the boundary, dots indicate countries and regions, blue dots show an increasing trend, and red dots indicate the opposite (Figure 5).

This figure shows the decomposition of disease metrics across Socio-Demographic Index (SDI) regions in the Global Burden of Disease (GBD) study. The main factors analyzed are aging (red), epidemiological changes (blue), and population growth (green). The four subplots depict changes in incidence (A), prevalence (B), mortality (C), and Disability-Adjusted Life Years (DALYs) (D). Incidence rates have risen significantly due to population growth, particularly in low and low-middle SDI regions. The effect of aging on incidence is minor in high and high-middle SDI regions. Prevalence changes are mainly driven by population growth, with all SDI regions showing an upward trend. Epidemiological changes particularly influence prevalence in low and low-middle SDI regions. Mortality changes are primarily driven by epidemiological shifts, especially in low and low-middle SDI regions, while population growth has a smaller effect on mortality in high SDI regions and globally. DALY changes are mainly driven by epidemiological factors, showing a decline across all SDI regions, especially in low and low-middle SDI areas. The impact of population growth on DALYs is minimal in high SDI regions. Overall, population growth is the primary driver of the increasing disease burden in various SDI regions, while epidemiological changes help reduce mortality and DALYs. Aging has a greater impact in high SDI regions but a relatively small effect globally. These findings suggest that low SDI regions should focus on strategies addressing population growth and epidemiological changes to mitigate disease burden (Figure 6).

ARIMA (Auto-Regressive Integrated Moving Average) and ES (Exponential Smoothing) are widely used time series forecasting methods. In the context of the Global Burden of Disease (GBD) database, they provide reliable predictions for trends in various health indicators. In this study, we applied both models to forecast the future trends of eight data points and compared their predictive performance. The ARIMA model outperformed the ES model in terms of smaller ME, RMSE, MAE, and MASE, indicating a better fit. For AIC and BIC values, the ARIMA model’s values were below 500, while the ES model’s values exceeded 500, further supporting the ARIMA model’s superiority. Residual checks showed that both models passed the white noise test. The ARIMA model’s residuals followed a normal distribution, with a Shapiro test p-value of less than 0.05, whereas the ES model’s p-value was greater than 0.05 (Supplementary Table 5). In summary, the ARIMA model proved to be superior. Regarding long-term trends, solid lines in the figures represent actual observed values, and dashed lines represent model-predicted future trends. In most cases, the forecasted trends suggest a gradual decline in disease burden indicators from 2020 (Figure 7).