Date of Award

1998

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Biometry and Epidemiology

College

College of Graduate Studies

First Advisor

Eberhard O. Voit

Second Advisor

Karen G. Burnett

Third Advisor

Rebecca G. Knapp

Fourth Advisor

G. Malcolm Meaburn

Fifth Advisor

Philip F. Rust

Abstract

Fish and shellfish are a beneficial food source due to their high protein content, low saturated fat content, and low cholesterol content; they possess beneficial omega-3 fatty acids and antioxidants such as vitamin E and selenium. Recent studies identified a decreased risk of sudden cardiac death in seafood consumers. Per capita seafood consumption increased 27% from 1975 to 1990, with nearly 95% of the U.S. population consuming some kind of fish or shellfish product each year. But seafood consumption may pose health risks which need to be weighed against benefits. One of potential risks is the toxicity of methylmercury. Mercury accumulates dramatically in the aquatic food web, where nearly all is converted into methylmercury. Almost all fish and shellfish contain at least trace amounts of methylmercury, and seafood consumption is the source of nearly all human methylmercury exposure. Major methylmercury poisonings provide evidence that the health effect of greatest concern from methylmercury exposure is neurotoxicity in the fetus and infant. Exposure assessments are the crucial step in identifying populations at risk of deleterious effects. Although several methods are currently available to project exposure, additional techniques may be necessary to accurately characterize exposure. One major complication in assessing exposure is the fact that all contributing factors are subject to variability, uncertainty, or to both. Variability describes heterogeneity in a well-characterized population, while uncertainty accounts for partial ignorance about a characteristic. Unfortunately, current exposure estimation techniques fail to simultaneously represent the full range and probability of exposures, account for dependencies among variables, and partition effects of variability from uncertainty. Thus, an innovative statistical methodology is developed here to overcome these deficiencies. Statistical techniques are also utilized to determine those input variables that are most critical to exposure. The different methods discussed and developed here may be used to estimate contaminant exposure from seafood consumption. Among other contaminants and marine species, methylmercury exposure from the consumption of Northern lobster, Homarus americanus, is currently of interest. Lobster is one of the more commonly consumed marine species, and it has the potential to bioaccumulate high methylmercury levels.

Rights

All rights reserved. Copyright is held by the author.

Download

Share

COinS