Abstract Details
Activity Number:
|
115
|
Type:
|
Topic Contributed
|
Date/Time:
|
Monday, August 5, 2013 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Section on Statistics in Epidemiology
|
Abstract - #307852 |
Title:
|
Varying-Coefficient Deming Regression (VCDR) and Its Application in Bone Densitometry
|
Author(s):
|
Ying Lu*+ and Chong Gu and Bo Fan and Selwyn Au and John A Shepherd
|
Companies:
|
VA Palo Alto Health Care System & Stanford University and Purdue University and University of California, San Francisco and VA Palo Alto Health Care System and University of California, San Francisco
|
Keywords:
|
Deming regression ;
Bone Mineral Density ;
Calibration ;
Non-linear regression ;
Varying Coefficients ;
NHANES
|
Abstract:
|
Deming regression (DR) has been successfully used to establish cross-calibration (CC) formulas for bone mineral densities (BMD) between manufacturers at several anatomic sites. In a recent study to establish CC for whole body BMD, DR was not used because the relationship between BMDs from Hologic and GE scanners varies with subject's weight, total fat and lean mass. Separate linear regressions were used for people with weight above or below 40 kg (Shepherd et al.,JBMR 2012), which don't meet the self-consistency requirements and discontinue at 40 kg. In this paper, we present a new varying-coefficient DR (VCDR) that allows the intercept and slope be non-linear functions of covariates. Estimation procedures and R programs were developed and applied to derive a consistent calibration formula that varied its coefficients based on covariates. Our results showed this VCDR effectively removed all systematic bias in Shepherd's formulas although the reduction of total calibration error was minimal. This formula can be used to translate the national normative whole body BMD data measured by Hologic scanners in the National Health and Nutrition Examination Study (NHANES) to GE scanners.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2013 program
|
2013 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Continuing Education program, please contact the Education Department.
The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Copyright © American Statistical Association.