New statistical approaches for adjusting seasonal variation in vitamin D concentration measurement, with application in the PLCO nested case-control study of serum vitamin D concentration and prostate cancer risk
Principal Investigator
About this CDAS Project
Study
PLCO
(Learn more about this study)
Project ID
2009-0006
Initial CDAS Request Approval
Nov 11, 2009
Title
New statistical approaches for adjusting seasonal variation in vitamin D concentration measurement, with application in the PLCO nested case-control study of serum vitamin D concentration and prostate cancer risk
Summary
Understanding the role of vitamin D in various diseases, including cancers, would have a tremendous public health impact. One of the major challenges in studying the relationship between vitamin D and a disease outcome is how to quantify a person's vitamin D level, as It is widely recognized that the vitamin D level like 25-hydroxyvitamin D [25(OH)D] in human body varies from season to season. The disease-vitamin D relationship can be obscure if such seasonal variation is not properly addressed. The locally weighted polynomial regression (LOESS) (LOESS), has been commonly used to estimate the seasonal trend. We propose two alternative statistical approaches for dealing with the seasonal variations, including one using the sine curve and the one using the penalized regression spline. To demonstrate the application of the proposed approaches, we plan to apply the new methods to reanalyze the recently published PLCO nested case-control study of serum vitamin D concentration and prostate cancer risk.
Aims
Use new proposed statistical approaches to reanalyze the recently published PLCO nested case-control study of serum vitamin D concentration and prostate cancer risk (Ahn et al . JNCI, 2008, 100: 796-804)
Collaborators
Jiyoung Ahn (New York University)