Abstract
INTRODUCTION: The objective of the study was to identify trends in incidence of adult diffuse gliomas in the United States and evaluate the contribution of age, period, and cohort effects to the trends. METHODS: Using the Surveillance, Epidemiology, and End Results 9 database, primary diffuse glioma patients (≥20 years old) diagnosed from 1973 to 2014 were identified. Incidence trends were analyzed using joinpoint regression and age-period-cohort modeling. RESULTS: Overall, the incidence for adult glioma decreased slowly from 1985 to 2014 (annual percent change [APC] = 0.5%, 95% confidence intervals [CI], 0.3%-0.6%). In histology subtype-stratified analysis, glioblastoma and nonglioblastoma exhibited opposite trends. The incidence for glioblastoma increased from 1978 to 2014 (APC for year 1978-1992 = 2.7%, 95% CI, 1.8%-3.6%; APC for 1992-2014 = 0.3%, 95% CI, 0%-0.6%), while the incidence for nonglioblastoma decreased significantly from 1982 to 2014 (APC = 2.2%, 95% CI, 2.0%-2.5%). Age-period-cohort modeling revealed significant period and cohort effects, with the patterns for glioblastoma and nonglioblastoma distinctive from each other. Compared with adults born 1890s, those born 1920s had approximately 4-fold the risk of glioblastoma after adjustment of age and period effects, while the risk of nonglioblastoma was reduced by half in individuals in the 1939 cohort as compared with those in the 1909 cohort. CONCLUSIONS: The results support the hypothesis of etiological heterogeneity of diffuse gliomas by histology subtypes. The established risk factors cannot fully explain the distinct patterns by histology subtypes, which necessitate further epidemiological studies.