Rising burden of early-onset lung cancer among Chinese adolescents and young adults (1990-2023): a Bayesian age-period-cohort analysis with projections to 2038.

BACKGROUND

Early-onset lung cancer (diagnosed at age 15-39 years) remains uncommon, yet incidence is rising in several world regions. China-the largest contributor to global lung cancer mortality-has not had its youth burden comprehensively characterized or projected.

METHODS

Using the Global Burden of Disease Study 2023 (GBD 2023), we extracted annual incidence, prevalence, mortality, and disability-adjusted life-years (DALYs) for Chinese adolescents and young adults (AYA) aged 15-39 years from 1990 to 2023. Joinpoint regression quantified temporal slopes; age-period-cohort (APC) and Bayesian APC (BAPC) models dissected historical drivers. Two complementary forecasting frameworks, BAPC (structural) and autoregressive integrated moving average (ARIMA; time-series) were used to generate burden trajectories to 2038.

RESULTS

Between 1990 and approximately 2020, the age-standardized disease burden generally declined, followed by a recent increase through 2023. Males consistently exhibit a higher disease burden than females; the 35-39-year age band showed the heaviest burden. Age-period-cohort analysis revealed divergent cohort effects: Cohort rate ratios generally declined across successive birth cohorts, with lower risks observed among cohorts born after 1985. For both sexes combined, BAPC point projections indicated that ASIR would increase from 1.53 to 2.54 per 100,000 and the age-standardized DALY rate from 64.71 to 79.77 per 100,000 between 2023 and 2038. These endpoint predictions corresponded to post hoc annualized geometric changes of +3.44% (UI - 9.37 to +7.83%) and +1.41% (UI - 7.86 to +5.31%), respectively. These estimates describe changes in age-standardized rates rather than absolute incident cases or DALY counts and are not BAPC drift coefficients. In contrast, ARIMA suggested modest declines or relatively flat trajectories for ASIR and ASMR, with wide prediction intervals. The discrepancy reflects model philosophy-BAPC extrapolates age-period-cohort structures, whereas ARIMA extrapolates short-term autoregressive patterns.

CONCLUSION

After decades of decline, lung cancer burden in Chinese adolescents and young adults is creeping upward and may accelerate. Rising computed tomography screening intensity, persistent ambient pollution, evolving tobacco-product use, and hereditary susceptibility warrant urgent scrutiny. These findings inform precision prevention strategies-targeted tobacco control, cleaner-air policies, and risk-stratified early detection-to avert a foreseeable surge in early-onset lung cancer in China.