Date of Award

2004

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Biometry and Epidemiology

College

College of Graduate Studies

First Advisor

David G. Hoel

Second Advisor

Daniel T. Lackland

Third Advisor

Lawrence C. Mohr

Fourth Advisor

Stuart R. Lipsitz

Fifth Advisor

Donald Frey

Abstract

The purpose of this dissertation was to gain a better understanding of the possible association between childhood leukemia and nuclear facilities by conducting a meta-analysis and by the use of spatial and temporal models with respect to Pickering Nuclear Generator (PNG) in Ontario, Canada. The meta-analysis was conducted to combine and statistically analyze the many studies of childhood leukemia in the vicinity of nuclear facilities. Our focus was on studies that calculated SMRs or SIRs for individual nuclear sites. Due to variability between studies in defining age and geographic zones, eight separate analyses were performed based on age and zone stratification levels. One hundred and forty-six sites were used in at least one analysis. Unadjusted models, fixed effects models, and random effects models were used for each of the eight analyses. Meta-rates greater than one were found in all models at all stratification levels. Further, statistical significance at 95% confidence intervals was often achieved. Within geographic zones (as established by the meta-analysis), the 0-9 age group experienced higher rates than the 0-25 age group. There does not appear to be publication bias in the meta-analysis. To better understand the temporal and spatial relationship between radiation from PNG and childhood leukemia, smoothed moving rates through time and a spatial model (Score Test of Lawson and Waller) were used that allowed for a more comprehensive description of disease patterns. No apparent relationship between childhood leukemia and PNG was detected. In the temporal analysis, moving SIRs remained near one for the entire time-period for the census subdivisions of Pickering and Ajax (which contain PNG). Zones based on distance from PNG were created for the spatial analysis. The highest rates were found in the innermost and outermost zones, with the highest population in the outer zones. No significant results were found with the Score Test of Lawson and Waller.

Rights

All rights reserved. Copyright is held by the author.

Download

Share

COinS