In the region of Puglia, Italy, the mortality rates from primary liver cancer (PLC) show a considerable geographical variability. In an area including the city of Bari and the northern province [high-risk (HR) area] the mortality rates are significantly higher than elsewhere in the region [low-risk (LR) area]. The aim of this study is to analyze mortality because of PLC from 1980 to 2002 in the HR and LR areas using an age-period-cohort model to identify their respective trend characteristics and the differences that emerge from their comparison. Attention was focused on the identifiable effects, which are on the drift and on the deviations from this regular trend (curvature), specifically attributable to the cohort and the period effects by applying the method recently proposed by Carstensen. The HR area is characterized by a more marked increase in the mortality trend compared with that observed in the LR area, as indicated in the model by a greater drift effect. In both areas the cohort (curvature) effect shows a decreasing trend starting from the early 1930s whereas the period effect shows a peak in the first half of the 1990s. Despite the correspondence of the trends, the rate ratio of death from PLC between the two areas is not constant and tends to rise uniformly by age, birth cohort, and period of time: in 2002, the rate ratio of death was 1.40 [confidence interval (CI): 1.15-1.70] at 40 years old, 1.73 (CI: 1.55-1.93) at 60 years old, and 2.14 (CI: 1.92-2.38) at 80 years old.
Mortality trends for primary liver cancer in Puglia, Italy
MARTINELLI, DOMENICO;PRATO, ROSA
2010-01-01
Abstract
In the region of Puglia, Italy, the mortality rates from primary liver cancer (PLC) show a considerable geographical variability. In an area including the city of Bari and the northern province [high-risk (HR) area] the mortality rates are significantly higher than elsewhere in the region [low-risk (LR) area]. The aim of this study is to analyze mortality because of PLC from 1980 to 2002 in the HR and LR areas using an age-period-cohort model to identify their respective trend characteristics and the differences that emerge from their comparison. Attention was focused on the identifiable effects, which are on the drift and on the deviations from this regular trend (curvature), specifically attributable to the cohort and the period effects by applying the method recently proposed by Carstensen. The HR area is characterized by a more marked increase in the mortality trend compared with that observed in the LR area, as indicated in the model by a greater drift effect. In both areas the cohort (curvature) effect shows a decreasing trend starting from the early 1930s whereas the period effect shows a peak in the first half of the 1990s. Despite the correspondence of the trends, the rate ratio of death from PLC between the two areas is not constant and tends to rise uniformly by age, birth cohort, and period of time: in 2002, the rate ratio of death was 1.40 [confidence interval (CI): 1.15-1.70] at 40 years old, 1.73 (CI: 1.55-1.93) at 60 years old, and 2.14 (CI: 1.92-2.38) at 80 years old.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.