|
Activity Number:
|
25
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 3, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #301981 |
|
Title:
|
Estimation for Age-Period-Cohort Models: With Application to Mesothelioma Data in Alberta
|
|
Author(s):
|
Bei Jiang*+ and K. C. Carriere
|
|
Companies:
|
University of Alberta and University of Alberta
|
|
Address:
|
632 Central Academic Building, Edmonton, AB, T6G 2G1, Canada
|
|
Keywords:
|
Age-Period-Cohort Model ; Log-linear model ; Estimable functions ; Regression splines ; Smoothing splines ; Generalized Additive Models
|
|
Abstract:
|
Age-Period-Cohort (APC) models have been widely used by epidemiologist to analyze the trends of disease incidence and mortality. The linear dependency: "Cohort=Period-Age" results in the well-known "nonidentifiability" problem. We review the "solutions" built on estimable functions (Holford, 1983). We show how the restricted regression splines can be used to extend the classical framework of factorial APC models (Heuer 1997). Further, we consider the use of Generalized Additive Models (GAM) for the estimation and present how the "non-identifiability" problem can be dealt with accordingly. Finally, these methods are illustrated to the Mesothelioma data in Alberta.
|
