Global, regional, and national burden of ovarian cancer, 1990-2021, and projections to 2050: a cross-sectional analysis of the Global Burden of Disease Study 2021.

Ovarian cancer ranks as the eighth most common cancer and is the fifth leading cause of cancer-related deaths among women globally [ 1 , 2 ] . Approximately 314 000 women worldwide are diagnosed with ovarian cancer each year, and 207 000 women die from the disease [ 3 ] . According to recent estimates, ovarian cancer accounts for a substantial economic burden on healthcare systems, affecting both the patients and society at large [ 4 , 5 ] . The complexity of this disease is underscored by its heterogeneous nature; various histological types exhibit distinct biological behaviors and responses to treatment, complicating diagnosis and management strategies [ 6 ] . In clinical practice, transvaginal ultrasound is widely recognized as the first-line imaging modality of choice for evaluating adnexal pathology [ 7 – 9 ] . Numerous efforts have been undertaken to develop more objective ultrasound-based methods for distinguishing between benign and malignant adnexal tumors. One such approach is the Risk of Malignancy Index (RMI), a scoring system that integrates menopausal status, transvaginal ultrasound findings, and serum cancer antigen 125 levels [ 10 ] . Many studies have validated the diagnostic efficacy of the RMI in classifying adnexal masses [ 9 , 11 – 14 ] . For patients with suspected malignant ovarian tumors, a computed tomography (CT) scan encompassing the chest, abdomen, and pelvis is mandatory prior to planned surgery. This imaging is critical for excluding secondary malignancies, thromboembolic events, and multifocal intraparenchymal distant metastases that could render the tumor unresectable [ 15 ] . While CT remains the most commonly utilized modality for staging ovarian cancer, magnetic resonance imaging and positron emission tomography-CT are increasingly employed in specialized centers for staging advanced cases. These functional imaging techniques can provide supplementary information to assess the feasibility of complete cytoreductive surgery and inform subsequent treatment strategies [ 16 ] . There are significant differences in the incidence and mortality rate of ovarian cancer among various regions and countries [ 17 ] . Some regions may suffer from inadequate medical resources and healthcare services, which causes delays in cancer screening and treatment, ultimately increasing the incidence and mortality rates of this disease. It is reported that the incidence of ovarian cancer is high in developed European countries and low in African countries [ 18 ] . All these heterogeneities make ovarian cancer a complicated disease and a global public health concern. Therefore, thorough research and analysis of ovarian cancer burdens in various regions and countries are crucial for creating more focused prevention and control strategies. The Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) included 100 983 data sources from 204 countries and territories and produced estimates of global health and health loss. This provides a unique opportunity to understand the trends in disease burden over the past three decades [ 19 ] . In this study, we aimed to report the global, regional, and national burden and trends associated with ovarian cancer and the burden attributed to risk factors, from 1990 to 2021, as well as projections up to 2050.

The dataset for this cross-sectional study is derived from the publicly available database of the Global Burden of Disease Study 2021 (GBD 2021), which includes epidemiological data on 371 diseases and injuries across 204 countries and regions from 1990 to 2021 [ 19 ] . The data were accessed through the official GBD visualization platform ( https://vizhub.healthdata.org/gbd-results/ ), with the following dimensions: Geographic Dimension: The data cover global-level statistics, categorized by five Social Demographic Index (SDI) levels (low, low-middle, middle, high-middle, and high SDI), 21 regions defined by the GBD framework, and 204 specific countries and territories [ 20 , 21 ] . Demographic Dimension: This study analyzes data across all age groups (the GBD database includes populations aged 15 to 95+), with age groups segmented in 5-year intervals. Core Indicators: The study extracted data on ovarian cancer incidence, prevalence, mortality, and disability-adjusted life years (DALYs), along with their corresponding age-standardized rates (ASRs): age-standardized incidence rate (ASIR), age-standardized prevalence rate (ASPR), age-standardized mortality rate (ASMR), and age-standardized DALY rate (ASDR). The corresponding 95% uncertainty intervals (UI) were also obtained. To assess the temporal changes in the burden of ovarian cancer from 1990 to 2021, this study employed two statistical methods: the estimated annual percentage change (EAPC) and Joinpoint regression analysis [ 22 , 23 ] . The EAPC was calculated using a log-linear regression model [ln( Y ) = α + β X + ε], with the formula 100 × [exp(β) − 1], where β represents the regression coefficient, X denotes the year, and Y is the natural logarithm of the age-standardized rate. The 95% confidence interval (CI) of EAPC was used to assess the statistical significance of trends. An EAPC greater than zero indicates an upward trend, while a value less than zero indicates a downward trend. The Joinpoint regression analysis was performed using the Joinpoint Regression Program 4.9.1.0 software, which identifies inflection points in the time trend using the Monte Carlo permutation test method, and calculates the average annual percentage change (AAPC) to describe the overall trend of the disease burden, capturing non-linear characteristics of the trend and pinpointing the specific time points at which trend changes occur. The study employed the APC model framework to decompose and analyze the three major factors influencing the changes in disease burden: age effect, period effect, and cohort effect. The age effect reflects the natural variations in disease risk over an individual’s life course; the period effect captures the short-term impact of external factors (such as advancements in medical technology or public health interventions) on disease burden; and the cohort effect reveals long-term differences in disease burden among different birth cohorts exposed to specific risk factors. The model was implemented using R software (version 4.3.3), with orthogonal decomposition applied to separate linear and non-linear components, and parameters estimated using weighted least squares. Model goodness-of-fit was assessed using the Wald χ 2 test [ 24 ] . This study employed a demographic decomposition method, which divides the changes in disease burden into three core components: changes in age structure, changes in population size, and changes in disease prevalence. By analyzing the relative contributions of each factor, the primary drivers of disease burden change are identified. The specific decomposition method and calculation formulas can be found in the references [ 25 , 26 ] . A BAPC model was applied to forecast the trends in ovarian cancer disease burden from 2022 to 2050. Compared to traditional models, the BAPC model has the advantage of using second-order random walk priors to smooth the age, period, and cohort effects, effectively avoiding overfitting. The model was computed using the integrated nested Laplace approximation method, which efficiently obtained the marginal posterior distribution, thereby avoiding the use of traditional Markov Chain Monte Carlo methods. The Concentration Index of Inequality (CII) was calculated by matching the cumulative proportion of DALYs with the cumulative population distribution sorted by SDI, followed by numerical integration of the area under the Lorenz curve [ 27 ] . Inequality was measured using the Slope Index of Inequality (SII) and the Concentration Index of Inequality (CII) to assess both absolute and relative health inequality in disease burden. The SII was calculated by regressing the DALYs rate against the SDI, using the midpoint values of the cumulative population distribution sorted by SDI. The CII was calculated by matching the cumulative proportion of DALYs with the cumulative population distribution sorted by SDI, followed by numerical integration of the area under the Lorenz curve [ 28 ] . Additionally, frontier analysis was performed, constructing a frontier model based on ASDR and SDI. This model aims to identify the theoretical minimum ASDR that each country or region could achieve at a given level of development. By quantifying the gap between the actual disease burden and the potential minimum burden, areas for improvement can be pinpointed. The study employed local weighted regression combined with local polynomial regression, using varying smoothing spans to capture the non-linear relationship between SDI and ASDR. To ensure the robustness of the analysis, 100 bootstrap resampling iterations were performed, and the average ASDR for each SDI value was computed [ 28 ] . Spearman’s rank correlation analysis was conducted to examine the relationship between SDI and ASR. To control for the false positive rate arising from multiple comparisons, the Benjamini–Hochberg method was used to adjust the false discovery rate (<0.05). Furthermore, local weighted scatterplot smoothing was applied to fit non-linear trends, providing further insight into the complex relationship between SDI and disease burden. All data cleaning, organization, and statistical analyses were performed in the R 4.3.3 software (2024) environment. Data processing and visualization were primarily carried out using the dplyr and ggplot2 R packages, with report generation facilitated by the officer package. All statistical tests were two-sided, with a significance level set at 0.05. This cross-sectional study has been reported in line with the STROCSS guidelines [ 29 ] .

In 2021, the ASIR of ovarian cancer was 6.71 (95% UIs: 6.07, 7.28) per 100 000, the ASPR was 28.08 (95% UI: 25.26, 30.64) per 100 000, the ASMR was 4.06 (95% UI: 3.67, 4.40) per 100 000, and the ASDR was 115.15 (95% UI: 104.58, 125.21) per 100 000 years. These figures indicate that in 2021, there were 298 876 new cases of ovarian cancer (95% UI: 270 729.82, 324 501.02), 1 222 425.26 existing cases (95% UI: 1 102 105.85, 1 332 111.98), 185 608.68 deaths (95% UI: 167 961.98, 201 012.67) due to ovarian cancer, and 5 163 256.30 DALYs (95% UI: 4 692 422.55, 5 608 304.11) attributed to the disease. From 1990 to 2021, ovarian cancer’s ASIR, ASMR, and ASDR showed significant declines (with EAPCs and CIs all being negative), whereas the ASPR significantly increased (with EAPCs and 95% CIs all being positive) (Table 1 and Supplementary Digital Content Fig. 1, available at: http://links.lww.com/JS9/F28 ). Table 1 ASR and EAPC for global, 5 SDI regions, and 21 GBD regions in 1990 and 2021 ASIR ASPR ASMR ASDR 1990 (per 100 000 population, 95% UI) 2021 (per 100 000 population, 95% UI) EAPCs (95% CI) 1990 (per 100 000 population, 95% UI) 2021 (per 100 000 population, 95% UI) EAPCs (95% CI) 1990 (per 100 000 population, 95% UI) 2021 (per 100 000 population, 95% UI) EAPCs (95% CI) 1990 (per 100 000 population, 95% UI) 2021 (per 100 000 population, 95% UI) EAPCs (95% CI) Global 7.22 (6.65, 7.87) 6.71 (6.07, 7.28) −0.38 (−0.43, −0.32) 27.62 (25.26, 30.26) 28.08 (25.26, 30.64) 1.17 (0.94, 1.39) 4.73 (4.38, 5.12) 4.06 (3.67, 4.40) −0.62 (−0.68, −0.57) 132.48 (121.34, 145.63) 115.15 (104.58, 125.21) −0.59 (−0.64, −0.54) SDI level  High SDI 11.96 (11.45, 12.34) 8.40 (7.84, 8.81) −1.21 (−1.33, −1.09) 50.02 (48.43, 51.41) 37.89 (35.88, 39.41) 1.24 (1.16, 1.31) 7.54 (7.15, 7.82) 5.08 (4.62, 5.37) −1.34 (−1.44, −1.25) 208.23 (200.58, 214.57) 133.80 (125.32, 140.06) −1.48 (−1.57, −1.39)  High-middle SDI 7.87 (7.13, 8.49) 6.93 (6.10, 7.76) −0.53 (−0.62, −0.45) 31.64 (28.23, 34.39) 30.37 (26.64, 34.23) 1.38 (1.23, 1.53) 4.95 (4.54, 5.29) 4.16 (3.68, 4.63) −0.70 (−0.80, −0.60) 149.55 (135.03, 161.63) 119.79 (105.13, 134.14) −0.87 (−0.96, −0.79)  Middle SDI 4.33 (3.73, 5.12) 5.81 (5.07, 6.50) 0.73 (0.65, 0.81) 16.34 (13.55, 19.54) 25.43 (21.73, 28.62) 1.55 (1.44, 1.67) 2.67 (2.35, 3.12) 3.18 (2.78, 3.57) 0.32 (0.23, 0.41) 82.58 (71.58, 97.89) 97.76 (85.77, 109.48) 0.31 (0.22, 0.40)  Low-middle SDI 3.69 (2.90, 5.02) 6.01 (5.22, 7.07) 1.51 (1.45, 1.58) 12.57 (9.87, 16.91) 22.92 (19.71, 26.94) 1.53 (1.42, 1.64) 2.43 (1.93, 3.25) 3.72 (3.24, 4.40) 1.34 (1.27, 1.40) 73.74 (58.18, 99.58) 111.40 (96.54, 131.20) 1.29 (1.23, 1.34)  Low SDI 3.67 (2.47, 5.24) 5.42 (4.12, 6.43) 1.18 (1.09, 1.28) 11.23 (7.57, 16.29) 18.42 (13.85, 22.07) 1.49 (1.37, 1.60) 2.65 (1.78, 3.76) 3.74 (2.87, 4.43) 1.06 (0.97, 1.15) 78.39 (52.96, 112.38) 108.72 (82.67, 129.40) 0.99 (0.89, 1.08) 21 regions  Andean Latin America 3.65 (2.93, 4.78) 6.90 (5.27, 8.68) 2.12 (1.72, 2.52) 12.81 (10.21, 17.00) 28.51 (21.49, 36.37) −0.07 (−0.13, 0.00) 2.48 (1.99, 3.22) 4.22 (3.27, 5.26) 1.77 (1.41, 2.13) 73.36 (58.97, 96.23) 125.60 (95.71, 158.01) 1.80 (1.43, 2.17)  Australasia 14.75 (13.63, 15.69) 7.18 (6.51, 7.81) −2.21 (−2.58, −1.84) 63.19 (58.45, 67.04) 32.16 (29.64, 34.60) 0.09 (−0.11, 0.29) 9.64 (8.89, 10.26) 4.68 (4.12, 5.08) −2.23 (−2.57, −1.88) 261.75 (242.26, 276.49) 116.72 (106.15, 125.99) −2.50 (−2.83, −2.16)  Caribbean 4.78 (4.37, 5.88) 6.66 (5.70, 8.07) 0.91 (0.73, 1.09) 17.58 (16.06, 21.45) 26.86 (22.83, 32.64) −0.14 (−0.31, 0.03) 3.17 (2.91, 3.85) 4.17 (3.59, 4.95) 0.79 (0.63, 0.94) 93.47 (84.91, 116.94) 124.35 (106.00, 151.66) 0.80 (0.65, 0.96)  Central Asia 4.28 (3.90, 4.68) 6.10 (5.31, 6.94) 1.22 (1.06, 1.38) 16.40 (14.97, 18.00) 24.01 (20.78, 27.37) 0.18 (−0.02, 0.38) 2.86 (2.58, 3.15) 4.08 (3.57, 4.63) 1.25 (1.10, 1.41) 90.99 (82.60, 99.60) 124.11 (108.08, 141.31) 1.05 (0.90, 1.21)  Central Europe 10.92 (10.54, 11.30) 10.80 (9.93, 11.68) 0.00 (−0.17, 0.17) 43.14 (41.51, 44.59) 43.47 (39.98, 47.08) −0.22 (−0.29, −0.15) 7.22 (6.96, 7.48) 7.40 (6.79, 7.96) 0.11 (−0.05, 0.27) 221.23 (213.40, 228.69) 205.22 (189.37, 221.33) −0.20 (−0.36, −0.04)  Central Latin America 5.35 (5.19, 5.51) 8.23 (7.16, 9.29) 1.38 (1.29, 1.47) 19.61 (19.09, 20.13) 34.93 (30.06, 39.55) 0.37 (0.25, 0.48) 3.54 (3.42, 3.64) 4.90 (4.32, 5.50) 1.05 (0.96, 1.13) 104.05 (101.06, 106.84) 150.76 (131.25, 170.19) 1.19 (1.11, 1.26)  Central Sub-Saharan Africa 2.83 (1.87, 4.32) 4.38 (2.49, 6.25) 1.49 (1.28, 1.70) 8.49 (5.61, 12.93) 14.60 (8.47, 20.94) 0.55 (0.31, 0.79) 2.09 (1.38, 3.15) 3.10 (1.69, 4.42) 1.36 (1.17, 1.56) 60.64 (40.14, 93.43) 90.01 (50.59, 129.58) 1.37 (1.17, 1.57)  East Asia 4.13 (3.06, 5.34) 4.21 (3.15, 5.46) −0.35 (−0.49, −0.20) 15.98 (11.05, 20.93) 19.58 (14.56, 25.49) 0.59 (0.48, 0.69) 2.61 (2.02, 3.31) 2.33 (1.72, 3.01) −0.90 (−1.09, −0.72) 82.65 (61.08, 107.10) 72.07 (53.13, 94.15) −0.92 (−1.09, −0.74)  Eastern Europe 10.30 (9.83, 10.75) 9.83 (8.77, 11.03) −0.22 (−0.40, −0.04) 42.79 (40.76, 44.80) 41.71 (36.90, 47.34) −0.94 (−1.10, −0.78) 6.25 (5.96, 6.52) 6.00 (5.37, 6.73) −0.21 (−0.39, −0.02) 200.42 (191.08, 209.58) 180.89 (160.79, 203.55) −0.44 (−0.62, −0.25)  Eastern Sub-Saharan Africa 5.23 (3.16, 7.47) 7.48 (5.31, 9.25) 1.08 (0.99, 1.17) 15.78 (9.41, 22.62) 25.41 (17.53, 31.71) −1.13 (−1.27, −0.98) 3.84 (2.33, 5.36) 5.21 (3.74, 6.35) 0.95 (0.87, 1.03) 111.66 (66.58, 160.61) 150.96 (106.18, 185.66) 0.91 (0.83, 0.99)  High-income Asia Pacific 6.28 (6.01, 6.54) 6.36 (5.69, 6.75) −0.01 (−0.16, 0.15) 27.88 (26.60, 29.05) 33.22 (30.01, 35.16) 1.30 (1.14, 1.47) 3.75 (3.55, 3.90) 3.26 (2.87, 3.50) −0.53 (−0.61, −0.45) 118.37 (114.08, 122.28) 98.97 (89.26, 104.57) −0.63 (−0.70, −0.55)  High-income North America 13.09 (12.52, 13.44) 8.88 (8.29, 9.27) −1.33 (−1.51, −1.15) 56.20 (54.51, 57.41) 38.38 (36.76, 39.93) −1.30 (−1.49, −1.11) 7.95 (7.50, 8.21) 5.44 (4.96, 5.71) −1.28 (−1.47, −1.10) 216.72 (208.63, 222.32) 139.13 (131.42, 145.12) −1.49 (−1.65, −1.33)  North Africa and Middle East 3.43 (2.72, 4.84) 4.87 (3.96, 5.63) 1.26 (1.19, 1.32) 12.13 (9.67, 17.44) 19.50 (15.67, 22.61) 1.63 (1.55, 1.70) 2.34 (1.88, 3.26) 3.14 (2.54, 3.65) 1.10 (1.03, 1.18) 68.09 (54.20, 96.89) 88.12 (71.92, 102.25) 0.95 (0.88, 1.01)  Oceania 2.73 (1.89, 3.77) 3.90 (2.32, 5.28) 1.14 (1.01, 1.26) 9.89 (6.67, 13.64) 15.47 (8.78, 21.50) 1.77 (1.68, 1.86) 1.49 (1.03, 2.00) 1.97 (1.23, 2.64) 0.99 (0.89, 1.08) 43.87 (29.72, 60.65) 59.04 (35.09, 80.86) 1.00 (0.90, 1.09)  South Asia 3.77 (2.80, 4.87) 5.99 (5.13, 7.14) 1.37 (1.27, 1.46) 12.30 (9.21, 16.05) 22.03 (18.49, 26.24) 1.83 (1.69, 1.97) 2.63 (1.95, 3.41) 3.92 (3.39, 4.67) 1.16 (1.07, 1.26) 79.51 (59.99, 102.65) 116.00 (99.82, 138.16) 1.09 (1.00, 1.18)  Southeast Asia 5.85 (4.63, 8.13) 9.28 (6.97, 12.04) 1.39 (1.27, 1.50) 23.71 (18.34, 33.88) 43.52 (32.08, 56.58) 1.84 (1.58, 2.10) 2.84 (2.32, 3.85) 3.96 (3.07, 5.12) 1.01 (0.93, 1.09) 91.18 (73.07, 126.59) 126.63 (96.38, 165.18) 0.99 (0.91, 1.07)  Southern Latin America 8.71 (8.00, 9.49) 7.69 (7.14, 8.23) −0.25 (−0.43, −0.08) 31.39 (28.75, 34.37) 31.73 (29.30, 33.94) 1.84 (1.72, 1.95) 5.98 (5.46, 6.54) 4.82 (4.44, 5.17) −0.51 (−0.68, −0.35) 169.10 (155.56, 184.70) 137.30 (127.41, 146.65) −0.51 (−0.67, −0.36)  Southern Sub-Saharan Africa 4.98 (4.05, 6.25) 7.71 (5.88, 8.93) 1.61 (1.47, 1.74) 17.05 (14.23, 20.87) 26.13 (20.12, 30.57) 1.88 (1.81, 1.94) 3.45 (2.64, 4.39) 5.38 (4.05, 6.16) 1.63 (1.50, 1.77) 98.60 (79.59, 124.62) 150.23 (114.98, 174.11) 1.59 (1.45, 1.73)  Tropical Latin America 5.56 (5.34, 5.75) 6.43 (6.05, 6.77) 0.20 (0.09, 0.32) 20.22 (19.48, 20.86) 26.41 (24.97, 27.71) 1.89 (1.78, 1.99) 3.70 (3.50, 3.84) 3.97 (3.67, 4.20) −0.00 (−0.12, 0.12) 109.27 (105.00, 113.00) 117.39 (110.28, 123.42) −0.02 (−0.13, 0.08)  Western Europe 13.41 (12.79, 13.88) 8.99 (8.35, 9.44) −1.35 (−1.44, −1.25) 57.40 (55.03, 59.39) 40.81 (38.95, 42.49) −2.04 (−2.46, −1.62) 8.31 (7.85, 8.65) 5.55 (5.03, 5.90) −1.35 (−1.42, −1.28) 227.51 (217.90, 236.02) 140.16 (131.08, 146.93) −1.61 (−1.68, −1.54)  Western Sub-Saharan Africa 2.25 (1.55, 2.83) 3.62 (2.40, 4.64) 1.59 (1.50, 1.68) 6.77 (4.67, 8.57) 11.78 (7.92, 15.22) 2.74 (2.29, 3.19) 1.66 (1.14, 2.10) 2.58 (1.69, 3.28) 1.48 (1.40, 1.56) 47.14 (32.69, 59.94) 72.64 (48.64, 93.39) 1.46 (1.37, 1.55) ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate; EAPC, estimated annual percentage change; GBD, Global Burden of Diseases, Injuries, and Risk Factors Study; SDI, Social Demographic Index. ASR and EAPC for global, 5 SDI regions, and 21 GBD regions in 1990 and 2021 ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate; EAPC, estimated annual percentage change; GBD, Global Burden of Diseases, Injuries, and Risk Factors Study; SDI, Social Demographic Index. Among the five SDI regions, the high SDI region had the highest rates in 2021, with an ASIR of 8.40 (95% UI: 7.84, 8.81) per 100 000, ASPR of 37.89 (95% UI: 35.88, 39.41) per 100 000, ASMR of 5.08 (95% UI: 4.62, 5.37) per 100 000, and ASDR of 133.80 (95% UI: 125.32, 140.06) per 100 000 years. In contrast, the low SDI region had the lowest ASIR [5.42 (95% UI: 4.12, 6.43) per 100 000] and ASPR [18.42 (95% UI: 13.85, 22.07) per 100 000], while the middle SDI region had the lowest ASMR [3.18 (95% UI: 2.78, 3.57) per 100 000] and ASDR [97.76 (95% UI: 85.77, 109.48) per 100 000 years]. From 1990 to 2021, ASIR, ASMR, and ASDR significantly declined in the high SDI and high-middle SDI regions, while all other regions saw significant increases in their respective ASRs (Table 1 and Supplementary Digital Content Fig. 1, available at: http://links.lww.com/JS9/F28 ). Among the 21 Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) regions, in 2021, Central Europe had the highest ASIR [10.80 (95% UI: 9.93, 11.68) per 100 000], ASMR [7.40 (95% UI: 6.79, 7.96) per 100 000], and ASDR [205.22 (95% UI: 189.37, 221.33) per 100 000 years], while Southeast Asia had the highest ASPR [43.52 (95% UI: 32.08, 56.58) per 100 000]. In contrast, Western Sub-Saharan Africa had the lowest ASIR [3.62 (95% UI: 2.40, 4.64) per 100 000] and ASPR [11.78 (95% UI: 7.92, 15.22) per 100 000], Oceania had the lowest ASMR [1.97 (95% UI: 1.23, 2.64) per 100 000 years], and ASDR [59.04 (95% UI: 35.09, 80.86) per 100 000 years]. From 1990 to 2021, Andean Latin America showed the most notable increases in ASIR, ASMR, ASDR, and Western Sub-Saharan Africa had the most significant increase in ASPR, while Australasia experienced the most notable decreases in ASIR, ASPR, ASMR, and ASDR (Table 1 and Supplementary Digital Content Fig. 1, available at: http://links.lww.com/JS9/F28 ). In 2021, the United Arab Emirates had the highest ASIR [25.40 (95% UI: 20.03, 32.42) per 100 000], ASMR [21.11 (95% UI: 16.64, 26.65) per 100 000], and ASDR [438.10 (95% UI: 342.36, 561.86) per 100 000 years], while the Republic of Seychelles had the highest ASPR [86.65 (95% UI: 63.63, 105.88) per 100 000]. In contrast, the Republic of Palau had the lowest ASIR [1.50 (95% UI: 1.09, 1.92) per 100 000], ASMR [0.73 (95% UI: 0.55, 0.93) per 100 000], ASDR [19.79 (95% UI: 14.32, 25.57) per 100 000], and the Republic of Mali had the lowest ASPR [5.51 (95% UI: 3.26, 7.54) per 100 000] (Fig. 1 ). From 1990 to 2021, Ecuador exhibited the most significant increase in all ASRs, while Australia’s ASIR, ASMR, ASDR, and Sweden’s ASPR showed the most notable declines (Supplementary Digital Content Fig. 2, available at: http://links.lww.com/JS9/F28 and Supplementary Digital Content Table 1, available at: http://links.lww.com/JS9/F29 ). Figure 1. ASR of ovarian cancer in 204 countries and regions for 2021 (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. ASR of ovarian cancer in 204 countries and regions for 2021 (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. Age trend analysis revealed that all ASRs for ovarian cancer increased with age in 2021. Specifically, the ASIR peaked in the 90–94 age group before declining, the ASPR peaked in the 60–64 age group before declining, the ASDR peaked in the 70–74 age group before declining, and the ASMR showed a continuous increase (Fig. 2 ). After excluding the effects of period and cohort, age effect analysis confirmed these trends (Supplementary Digital Content Fig. 3A–D, available at: http://links.lww.com/JS9/F28 ). Time trend analysis indicated that from 1990 to 2021, ASRs for ovarian cancer in all age groups globally, as well as in the high SDI and high-middle SDI regions, showed a slow decline, whereas in the middle SDI, low-middle SDI, and low SDI regions, ASRs showed a slow increase (Supplementary Digital Content Figs 4–7, available at: http://links.lww.com/JS9/F28 ). After excluding age and cohort effects, period effect analysis revealed a gradual decline in ASRs over time from 1990 to 2021 (Supplementary Digital Content Fig. 3E–H, available at: http://links.lww.com/JS9/F28 ). Additionally, cohort effect analysis indicated that later birth cohorts had lower ASRs than earlier birth cohorts (Supplementary Digital Content Fig. 3I–L, available at: http://links.lww.com/JS9/F28 ). Figure 2. Age–period analysis results of ovarian cancer ASR (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. Age–period analysis results of ovarian cancer ASR (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. Joinpoint segmented regression analysis demonstrated a significant downward trend from 1990 to 2021 for ASIR [average annual percentage change (AAPC) = −0.016 (95% CI: −0.017, −0.015)], ASMR [AAPC = −0.021 (95% CI: −0.022, −0.021)], and ASDR [AAPC = −0.558 (95% CI: −0.583, −0.534)], whereas the ASPR showed a significant increase [AAPC = 0.018 (95% CI: 0.011, 0.024)]. Specifically, both ASIR and ASMR showed inflection points in 1995, 2003, and 2015, with ASIR remaining unchanged since 2015 and ASMR showing a significant decline. ASPR and ASDR showed inflection points in 1995, 2002, and 2014, with ASPR significantly increasing since 2014 and ASDR significantly decreasing since 2014 (Fig. 3 and Supplementary Digital Content Table 2, available at: http://links.lww.com/JS9/F31 ). Figure 3. Join point analysis results of ovarian cancer ASR (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. Join point analysis results of ovarian cancer ASR (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. A significant correlation was found between ovarian cancer disease burden and the SDI (Spearman’s correlation coefficient P -values <0.05), with this correlation being consistent across the 21 GBD regions and 204 countries (Fig. 4 and Supplementary Digital Content Fig. 8, available at: http://links.lww.com/JS9/F28 ). ASRs exhibited an increasing and then decreasing trend as SDI increased. Specifically, ASRs began to decrease when SDI was approximately 0.8. The correlation coefficients and P -values are detailed in Supplementary Digital Content Table 3, available at: http://links.lww.com/JS9/F32 . Additionally, the EAPC for ASRs showed a similar significant correlation with SDI, with ASRs initially increasing and then decreasing as SDI increased. Specifically, ASRs began to decrease when SDI was approximately 0.6 (Supplementary Digital Content Fig. 9, available at: http://links.lww.com/JS9/F28 ). Figure 4. Correlation between ovarian cancer ASR and Social Development Index (SDI) across 21 regions (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. Correlation between ovarian cancer ASR and Social Development Index (SDI) across 21 regions (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. Health inequality analysis further revealed significant absolute and relative inequalities in ASDR with respect to SDI. The SII showed that the ASDR gap between countries with the highest and lowest SDI narrowed from 128.60 (95% UI: 102.36, 154.83) in 1990 to 56.13 (95% UI: 36.19, 76.07) in 2021 (Supplementary Digital Content Fig. 10, available at: http://links.lww.com/JS9/F28 ), indicating a reduction in absolute inequality, with the disease burden primarily concentrated in higher SDI countries. Simultaneously, the CII decreased from 0.24 (95% CI: 0.20, 0.27) in 1990 to 0.09 (95% CI: 0.05, 0.12) in 2021 (Supplementary Digital Content Fig. 11, available at: http://links.lww.com/JS9/F28 ), indicating a reduction in relative inequality between high and low SDI countries, with the disease burden primarily concentrated in higher SDI countries. Frontier analysis indicated that the 15 countries and regions with the largest gap from the optimal disease burden were the United Arab Emirates, Georgia, Bahamas, Latvia, Grenada, Poland, Bulgaria, Guyana, Greenland, Seychelles, Pakistan, Zimbabwe, Trinidad and Tobago, Lithuania, and Brunei Darussalam (potential improvement ranges: 190.17–422.81). In low SDI countries (SDI 0.85), the countries with the largest gap to the frontier were Lithuania, Ireland, the United Kingdom, Luxembourg, and Denmark (Supplementary Digital Content Fig. 12, available at: http://links.lww.com/JS9/F28 ). At the global level, both population growth and aging have contributed to the increase in the ASRs of ovarian cancer. Epidemiological changes have led to reductions in the ASIR, ASMR, and ASDR but have resulted in an increase in the ASPR. At the regional level, population growth has been a driving factor in the increase of ovarian cancer ASRs, while the effects of aging and epidemiological changes vary across regions. Specifically, aging has had the most significant impact on increasing ASRs in the middle SDI region, while it has led to the most notable reductions in ASIR, ASMR, and ASDR in Eastern Europe and reductions in ASPR in Western Europe. Aging has had the most pronounced effects on increasing ASIR and ASPR in the middle SDI region, as well as increasing ASMR and ASDR in the low-middle SDI region, while it has led to the most notable reduction in ASRs in the high SDI region (Supplementary Digital Content Fig. 13, available at: http://links.lww.com/JS9/F28 and Supplementary Digital Content Table 4, available at: http://links.lww.com/JS9/F33 ). From 2022 to 2050, the ASIR and ASPR of ovarian cancer are expected to continue rising, while ASMR and ASDR will initially decrease before increasing again (Fig. 5 ). By 2050, it is projected that the ASIR will reach 11.09 (95% UI: 9.60, 12.58) per 100 000, the ASPR will be 50.62 (95% UI: 41.54, 59.71) per 100 000, the ASMR will be 5.65 (95% UI: 5.12, 6.19) per 100 000, and the ASDR will be 169.13 (95% UI: 151.30, 186.97) per 100 000 years (detailed in Supplementary Digital Content Table 5, available at: http://links.lww.com/JS9/F34 ). Figure 5. Predictive analysis results of ovarian cancer ASR (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. Predictive analysis results of ovarian cancer ASR (A. ASIR; B. ASPR; C. ASMR; D. ASDR). ASDR, age-standardized disability-adjusted life year rate; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASPR, age-standardized prevalence rate; ASR, age-standardized rate. Currently, only two attributable risk factors – high body mass index (BMI) and occupational exposure to asbestos – have been reported. In 2021, 9.18% of ovarian cancer-related DALYs globally were attributable to high BMI, and 1.83% were attributable to occupational exposure to asbestos. The contributions of these risk factors to DALYs across the five SDI regions are similar (Supplementary Digital Content Fig. 14, available at: http://links.lww.com/JS9/F28 ). From 1990 to 2021, the ASDR attributable to high BMI increased globally and across all five SDI regions, while the ASDR attributable to occupational exposure to asbestos remained stable at lower levels (Supplementary Digital Content Fig. 15, available at: http://links.lww.com/JS9/F28 ).

Ovarian cancer poses a critical health challenge globally, being the leading cause of death from gynecologic cancer. The complexity of ovarian cancer is underscored by its heterogeneous nature, encompassing various histological types. Despite advancements in medical technology and treatment protocols, the prognosis remains poor for many patients diagnosed at advanced stages, thus highlighting the urgent need for enhanced understanding and strategic intervention in managing this malignancy [ 3 , 30 ] . A comprehensive investigation of ovarian cancer incidence, mortality, and DALYs is crucial for developing effective public health strategies. To provide reliable data for future public health policymakers and facilitate the development of targeted prevention and control strategies, we generated estimates of ovarian cancer prevalence and burden, examined the contribution of leading risk factors; and forecasted ovarian cancer prevalence in 2050. The results of this study indicated that in 2021, there were about 0.30 million new cases of ovarian cancer, 1.22 million existing cases, 0.19 million deaths due to ovarian cancer, and 5.16 million DALYs attributed to the disease. The global decline in ASIR, ASMR, and ASDR from 1990 to 2021 suggested improvements in early detection, treatment modalities, and survivorship. However, the concurrent rise in ASPR indicated prolonged survival among diagnosed individuals, reflecting advancements in therapeutic interventions. These trends align with global progress in cancer care but also emphasize the growing need for long-term management of ovarian cancer survivors, particularly in aging populations. In 2021, ASRs showed significant geographic variations, with the highest rates in the high SDI region and the lowest in the low SDI region. Regionally, the differences in trends were notable. High SDI regions showed declines in ASIR, ASMR, and ASDR. These declines may be linked to their advanced healthcare systems, higher education levels, and better access to preventive measures and treatments. In contrast, other regions experienced increases in their respective ASRs. This may be due to factors such as limited access to healthcare services, lack of awareness about early detection, and a higher prevalence of risk factors. Our results aligned with previous reports from GLOBOCAN [ 31 , 32 ] . For example, Cabasag et al [ 31 ] reported significant geographic variations in ovarian cancer incidence rates, with the highest rates in European countries with very high HDI and the lowest rates in African countries within the lowest HDI group. The APC analysis offers important insights into what influences the disease burden. The age effect showed that ASRs increased with age, which is consistent with the biological characteristics of ovarian cancer. Physiological changes over time place older women at a greater risk. The period effect, representing external influences, demonstrated a gradual decline in ASRs over the years. This decline may be due to advancements in public health initiatives, medical research, and the introduction of cancer-control programs. Over the past decade, significant efforts have been directed toward enhancing the prognosis of patients with ovarian cancer [ 33 – 37 ] . Notably, maintenance therapy using poly (ADP-ribose) polymerase (PARP) inhibitors – either as monotherapy or in combination with bevacizumab – has demonstrated substantial efficacy in newly diagnosed patients following platinum-based treatment [ 35 – 37 ] . BReast CAncer (BRCA) susceptibility gene mutations and homologous recombination deficiency (HRD) stand as the two most critical biomarkers for predicting responsiveness to PARP inhibitors [ 35 – 37 ] . In the PAOLA-1 trial, maintenance therapy combining olaparib and bevacizumab showed a trend toward improved overall survival in patients with BRCA mutations and HRD [ 37 , 38 ] . While immunotherapy has emerged as a groundbreaking treatment modality in numerous solid tumors, outcomes with immune checkpoint inhibitors in ovarian cancer have been underwhelming [ 39 ] . To date, immunotherapy – whether administered alone or in combination – has not been shown to improve overall survival in newly diagnosed, platinum-sensitive, or platinum-resistant ovarian cancer [ 39 ] . The definitive findings from ongoing trials, alongside novel innovative strategies, will help delineate the role of promising combination therapies in the management of ovarian cancer. The cohort effect showed that later birth cohorts experience lower ASRs, suggesting that changes in lifestyle, environmental factors, or improved access to healthcare across generations may contribute to a reduced disease burden. ASRs first increased and then declined as SDI increased, reaching a turning point at SDI = 0.8. This suggests that, in less developed regions, limited healthcare resources and preventive measures result in a higher burden of disease. As countries develop, they invest more in healthcare, education, and infrastructure, which helps reduce the disease burden. However, after reaching a certain level of development, factors like lifestyle changes, including rising obesity rates, may begin to negate the positive impacts of development on the ovarian cancer burden. The health inequality analysis further emphasized the disparities in disease burden. The demographic decomposition analysis revealed that population growth and aging contributed to the increase in ovarian cancer ASRs. In contrast, epidemiological changes had a mixed impact. Population growth created a larger pool of at-risk individuals, which increased the overall burden of ovarian cancer. Aging affected regions differently; in some areas, it significantly increased ASRs, while in others, it decreased them. Therefore, understanding these regional differences is crucial for developing targeted public health policies. Projections from 2022 to 2050 show that the ASIR and ASPR will likely keep rising, while the ASMR and ASDR are expected to drop initially before rising again. These projections underscore the necessity for ongoing initiatives in cancer prevention, early detection, and treatment. If the projected increases in incidence and prevalence occur, healthcare systems will face significant challenges in terms of resource allocation, including the need for more oncologists, treatment facilities, and support services. The later rise in ASMR and ASDR indicates a pressing need for new strategies to enhance long-term survival and the quality of life for ovarian cancer patients. Extensive research has been conducted on the risk factors for ovarian cancer. The most well-established ones include inherited predispositions (such as germline mutations in BRCA1/BRCA2 and Lynch syndrome), nulliparity, infertility, endometriosis, obesity, and age [ 40 – 42 ] . For individuals with inherited genetic susceptibilities to ovarian cancer, including mutations in BRCA1/BRCA2 or other homologous recombination genes, risk-reducing salpingo-oophorectomy (RRSO) is recommended as the most effective intervention to lower the risk of developing the disease [ 43 ] . Currently, RRSO stands as the standard of care for preventing ovarian cancer in women with a genetically elevated risk, offering an overall risk reduction of 75–95% [ 44 – 46 ] . In our study, we observed that high BMI and occupational exposure to asbestos were recognized risk factors for ovarian cancer. The rising impact of high BMI on the global disease burden is concerning, especially across all SDI regions. Given the rising prevalence of obesity worldwide, public health interventions should focus on promoting healthy lifestyles, including diet and physical activity, to reduce the risk of ovarian cancer. Although most countries have banned asbestos use, millions of workers are still exposed to it in factories. Every year, at least 90 000 people die from diseases related to asbestos exposure or cancer [ 47 ] . According to the World Health Organization’s Health Statistics database from 2009, Argentina, Brazil, Colombia, and Mexico reported the highest estimated ovarian cancer deaths linked to occupational asbestos exposure in the previous 5 years [ 48 ] . Although the contribution of occupational exposure to asbestos remains stable at a lower level, it underscores the need for strong workplace safety regulations to prevent carcinogen exposure. This study is the first to use an epidemiological model to analyze the global, regional, and national burden and trends of ovarian cancer, including the burden from risk factors from 1990 to 2021, and projections to 2050. Our results show a strong link between disease burden and socioeconomic development. This finding offers valuable insights for countries and healthcare professionals to create effective preventive measures, management policies, and treatment protocols. However, the study has several limitations. First, despite improvements in GBD analytical processes and methodologies, the quality of raw data and challenges in data collection still pose significant constraints to GBD analysis. Second, data availability in less developed regions is concerning, as a large number of undiagnosed cases of ovarian cancer may not accurately reflect the true disease burden. Third, ovarian cancer is a heterogeneous tumor, and because the GBD data do not provide histological information, we could not analyze its subtypes in detail. Finally, the analysis of attributable risk factors relies on the risk factors available in the GBD 2021 database, which might not include all clinically relevant factors. Therefore, we should interpret these results with caution. Future research should prioritize acquiring higher-quality epidemiological data and applying uniform methods and definitions to improve the accuracy and reliability of global ovarian cancer burden assessments.

This study provides a new basis for long-term trends in global ovarian cancer morbidity and mortality. While the global burden of ovarian cancer has decreased over the past 30 years, it is projected to rise in the next three decades. The burden of ovarian cancer varies significantly by age, time, and geographical location. Therefore, policymakers must urgently create effective prevention and control measures to reduce the burden of ovarian cancer, enhance family well-being, and lessen economic pressures on society.