Temporal changes in age at diagnosis and the proportion of early-onset cases among diagnosed cases of 16 selected cancers in the United States, 2000–2023

Abstract

Comprehensive evaluations of long-term changes in the age composition of diagnosed cancer cases, including the proportion of early-onset cases, remain limited. We conducted a descriptive registry-based analysis of diagnosed cancer cases aged ≥ 20 years across 16 selected cancer sites in the Surveillance, Epidemiology, and End Results (SEER)-21 database from 2000 to 2023. Mean and median age at diagnosis and age-group distributions were evaluated. Early-onset cases were defined as those diagnosed at ages 20–49 years, and the early-onset proportion was calculated by dividing the number of cases aged 20–49 years by the total number of diagnosed cancer cases aged ≥ 20 years. This within-case proportion did not represent an age-specific or age-standardized incidence rate in the underlying population. A total of 12,910,291 diagnosed cancer cases were included. The overall mean age at diagnosis decreased from 65.24 years in 2000 to 64.30 years in 2014 and subsequently increased to 65.40 years in 2023. The proportion of cases diagnosed at ages 20–49 years increased slightly from 13.17% in 2000 to a peak of 13.61% in 2005 and then declined to 11.55% in 2023. The absolute endpoint difference between 2014 and 2023 was − 0.77% points overall and + 3.75% points for colon and rectum cancer. Among diagnosed cases of the 16 selected cancers, the mean age at diagnosis was higher and the relative share of cases diagnosed at ages 20–49 years was lower in 2023 than in 2014 overall, although patterns varied by cancer site. Because the pooled overall estimates were composition-dependent, site-specific analyses provide a more informative basis for interpreting cancer-specific patterns. These findings describe changes in the age composition of diagnosed cancer cases and should not be interpreted as age-standardized incidence trends or population cancer burden.