Kundu · JCO global oncology 2026 · Global Burden of Disease ecological time-series and forecasting study · n=204 countries (GBD 2021 database)

Global Epidemiologic Trends of Early-Onset Cancers From 1990 to 2021 and Projection to 2040.

Cited 2 times in the scientific literature.

Level 4 - case-series / case-control

Global epidemiological modeling and descriptive time-series analysis (GBD 2021 dataset)

PubMed 41637674 · doi:10.1200/GO-25-00170 · record verified 2026-08-28

What was done

The authors evaluated global trends in early-onset cancer (EOC, defined as cancer diagnosed in individuals aged 15 to 49 years) from 1990 to 2021 and projected trends through 2040. Using data from the Global Burden of Disease (GBD) 2021 study covering 31 cancer types across 204 countries, they analyzed age-specific incidence rates (AsIR) and age-specific death rates (AsDR) per 100,000 patient-years. Projections up to 2040 were generated using autoregressive integrated moving average (ARIMA) and exponential smoothing state space models.

What was found

In 2021, there were approximately 23.6 million new EOC cases and 0.99 million deaths worldwide. Breast, nonmelanoma skin, and cervical cancers showed the highest AsIRs, whereas breast and lung cancers accounted for the highest AsDRs in 2021. Between 1990 and 2021, global AsIR increased whereas AsDR declined. Women, especially in high-income countries, were disproportionately affected. Obesity, tobacco use, and dietary patterns were identified as key risk factors. Modeling through 2040 projected relatively stable overall AsIR and continuing declines in AsDR.

Why it matters

This study tracks the expanding global burden of cancers diagnosed in adults under 50 years of age, demonstrating divergent trajectories between rising case incidence and falling mortality. It highlights target demographics and modifiable lifestyle factors to inform prevention and health system planning.

Limits

The study is based on modeled aggregate GBD estimates rather than primary clinical registries, which may obscure regional data quality variations. The abstract provides no precise effect sizes, rate numbers, or confidence intervals for temporal changes, nor does it quantify the proportional impact of specific risk factors. Projections to 2040 depend on historical assumptions that cannot anticipate future screening shifts or therapeutic breakthroughs.

Cited by