CNRS n° 6249 Chrono-Environment, Faculty of Medicine, 2, place Saint Jacques, 25030 Besançon cedex, France.
Cancer Epidemiol. 2011 Dec;35(6):521-5. doi: 10.1016/j.canep.2011.04.002. Epub 2011 May 28.
The long tenure of the Doubs cancer registry (France) and the late implementation of a mass screening program provide a unique opportunity to assess the relative contributions of age, period and cohort effects to the increase in female invasive breast cancer incidence, while avoiding the influence of an organized screening program.
Population and incidence data were provided for the Doubs region during the 1978-2003 period. Breast cancer counts and person-years were tabulated into 1-year classes by age and time period. Age-period-cohort models with parametric smooth functions were fitted to the data, assuming a Poisson distribution for the number of observed cases.
A total of 5688 incident cases of invasive breast cancer in women were diagnosed in women aged 30-84 years in the Doubs region between 1978 and 2003. The annual percentage increase in incidence is 2.09%. Age effects rise dramatically until age 50, and at a slower pace afterwards. Large cohort curvature effects (p<10(-6)), show departure from linear trends, with a significant peak for women born around 1940. Period curvature effects are lower in magnitude (p=0.01).
Both cohort and period effects are involved in the marked increase in breast cancer incidence over a 25-year period in the Doubs region. Although the future trend for breast cancer incidence is difficult to predict, the introduction of an organized screening program, and the sharp decline in hormone replacement therapy use will likely contribute to period effects in future analyses.
多布癌症登记处（法国）的长期运行和大规模筛查计划的后期实施为评估年龄、时期和队列效应对女性浸润性乳腺癌发病率上升的相对贡献提供了独特的机会，同时避免了筛查计划的影响。
为多布地区提供了 1978 年至 2003 年期间的人口和发病数据。乳腺癌病例和人年数据按年龄和时期划分为 1 年一组。使用参数平滑函数的年龄-时期-队列模型拟合数据，假设观察到的病例数呈泊松分布。
在 1978 年至 2003 年间，多布地区 30-84 岁的女性中诊断出 5688 例浸润性乳腺癌新发病例。发病率的年增长率为 2.09%。年龄效应在 50 岁之前急剧上升，之后以较慢的速度上升。大队列曲率效应（p<10(-6)）表明与线性趋势的背离，出生于 1940 年左右的女性出现显著高峰。时期曲率效应的幅度较小（p=0.01）。
在多布地区 25 年期间，乳腺癌发病率的显著上升涉及队列和时期效应。尽管乳腺癌发病率的未来趋势难以预测，但有组织的筛查计划的引入和激素替代疗法使用的急剧下降可能会对未来分析中的时期效应产生影响。
