GBD 2021: Global Endometriosis Burden, Trends (1990–2021) and Projections to 2036

Global estimates from GBD 2021 endometriosis burden report describe rising absolute burden alongside declining age-standardized rates from 1990 to 2021. The authors estimate that prevalent cases increased from 19.87 million (95% UI, 13.59–27.69 million) in 1990 to 22.28 million (95% UI, 15.52–30.41 million) in 2021. Over the same period, age-standardized prevalence fell from 757.10 to 556.98 per 100,000 (EAPC −1.02; 95% UI, −1.07 to −0.97), and age-standardized incidence fell from 119.65 to 88.52 per 100,000 (EAPC −1.00; 95% UI, −1.05 to −0.95). They characterize this pattern as reflecting demographic influences on absolute counts amid declining standardized rates.

Beyond prevalence and incidence, the authors report that disability-adjusted life-years (DALYs) attributable to endometriosis increased in absolute terms between 1990 and 2021. They highlight that the largest absolute DALY burden was concentrated in South Asia, East Asia, and Southeast Asia, while higher DALY age-standardized rates were observed in regions including Oceania and Eastern Europe (with additional high DALY age-standardized rates noted in parts of sub-Saharan Africa). They also note wider uncertainty intervals in low-SDI settings, which they describe as consistent with more limited primary data and greater reliance on modeled estimates. Overall, the reported results describe substantial geographic heterogeneity in both absolute burden and age-standardized rates.

By sociodemographic index (SDI), the authors describe middle-SDI and low-middle-SDI regions as having relatively higher prevalence and incidence in 2021, while high-SDI regions showed the lowest levels, alongside declining age-standardized trends across SDI categories over time. In their inequality analyses, they report declining absolute inequalities (slope index of inequality, SII) for prevalence, incidence, and DALY from 1990 to 2021, while relative inequalities (concentration index, CIX) increased, which they describe as favoring wealthier populations and varying across SDI groupings. Their decomposition analysis attributes the observed absolute increases primarily to population growth; aging contributed to the increase in prevalence, but had minimal or slightly negative effects on incidence, alongside contributions from epidemiological change as presented. These distributional analyses describe shifting absolute and relative gradients during a period of declining age-standardized rates.

In age-pattern analyses for 2021, the highest prevalence was reported in the 25–29 age group, and the highest incidence in the 20–24 age group. The authors also describe DALY impacts as concentrating in early-to-mid reproductive ages, with the greatest impacts in the 20–24, 25–29, and 30–34 age bands. They further note that the 15–19 age group had the second-highest incidence but the lowest prevalence in 2021, consistent with differing age profiles for incidence versus prevalent disease. Overall, the reported age-specific burden clusters within the reproductive-age distribution.

For forecasting, the authors report Bayesian age–period–cohort (BAPC) projections through 2036 showing continued declines in age-standardized rates, including example point estimates for 2036 of an age-standardized prevalence rate of 499.63 per 100,000, an age-standardized incidence rate of about 80 per 100,000, and an age-standardized DALY rate of 45.38 per 100,000. They describe a Poisson likelihood for counts, second-order random walk (RW2) priors for period and cohort effects, and fitting via INLA using the BAPC R package. Limitations they state include reliance on GBD modeled estimates where primary data are sparse (with related uncertainty), inability to disaggregate clinical subtypes or symptomatic status, and projections generated at a global aggregate level without hierarchical modeling of cross-country dependence. They frame the forecasts as intended to support policymakers and planning discussions. In combination, the authors present model-based downward projections alongside explicit caveats about data granularity and uncertainty.

Key Takeaways:

• The authors report rising absolute prevalent cases from 1990 to 2021 alongside declining age-standardized prevalence and incidence rates over the same period.
• Inequality analyses are described as showing declining absolute inequalities (SII) but increasing relative inequalities (CIX), and decomposition attributes most absolute increases to population growth with additional roles for aging and epidemiological change.
• BAPC projections through 2036 are reported as declining for age-standardized prevalence, incidence, and DALY rates, and the authors emphasize that these projections are model-based and reported with uncertainty.