|
Activity Number:
|
477
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Health Policy Statistics
|
| Abstract - #308773 |
|
Title:
|
Statistical Methods for Constructing Canadian Growth Curves
|
|
Author(s):
|
Stephane Tremblay*+
|
|
Companies:
|
Statistics Canada
|
|
Address:
|
100 Tunney Driveway, Ottawa, ON, K1A 0T6, Canada
|
|
Keywords:
|
Growth curves ; BMI ; LMS method ; Sampling weights ; Design effect
|
|
Abstract:
|
Growth curves are used by Health Care providers to determine if the growth of a child or fetus (for example) is within normal boundaries. The current charts used in Canada for height, weight and body mass index (BMI) are based on United States data. Child growth curves can now be developed based on the most recent cross-sectional data available in Canada. The presentation will describe the LMS (Lambda-Mu-Sigma) method using a Box-Cox normal distribution and a cubic spline smoothing technique to estimate Canadian growth curves for children aged 2 to 18. Special attention will be put on how we used auxiliary information (from the US) to handle edge effects, how we handled extreme sampling weights and how we adjust for the design effects when estimating such curves. Criteria for the selection of the best model using statistical and graphical diagnostic tools will also presented.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |
