Faculty Publications


Exposure methodology and findings for dietary nitrate exposures in children of Transylvania, Romania

Document Type



Children, Exposure assessment, Nitrate exposure, Romania, Transylvania

Journal/Book/Conference Title

Journal of Exposure Analysis and Environmental Epidemiology





First Page


Last Page



Aims: The primary focus of this exposure assessment work involved developing an exposure model and determining a numerical point estimate of the amount of biologically relevant nitrate/nitrite exposure that occurred for each child in the study. This assessment was done in support of two epidemiological studies. The first study was an epidemiological cohort study (three cohorts based on nitrate/nitrite exposure) that explored the relationship between high nitrate/nitrite exposure and neuropsychological development. The second study was a nested case/control study (cases of methemoglobinemia versus disease-free controls) that sought to explore the relationship between MHG and various risk factors for the disease. Methods: This work uses both dietary survey and environmental sampling and modeling in order to develop two point estimates of nitrate exposure in milligrams per kilogram per day of nitrite (the biologically active form of the hemoglobin-oxidizing agent) for the first 6 months of the child's life (2-months-of-age and 6-months-of-age point estimates). Methodologies included proxy interviews of primary caregivers, review of existing medical and environmental sampling and analysis. Results: Exposure to nitrate-nitrogen (with final calculations converted to the biologically active form of the toxin, nitrite) was categorized as high, medium, and low as determined from the distribution of the data derived from final exposure calculations at both the 2 -months-of-age point estimate and at the 6-months-of-age point estimate. These tertiles correspond to ≥1.5 mg/kg/day nitrite-nitrogen for high-exposure individuals, <1.5-≥0.1 mg/kg/day for medium-exposure individuals, and <0.1 mg/kg/day for low-exposure individuals. Analyses illustrate that over 90% of the nitrate exposure occurred through the consumption of liquids (water) at the 2-months-of-age point estimate while at the 6-months-of-age point estimate, a 10-fold change in the amount of solid consumables occurred. Conclusions: Final exposure calculations were well differentiated into three tertiles based on a point estimate of average daily intake of nitrite in milligrams per kilogram body weight per day at roughly 2 and 6 months of age. These categories of exposure, based on the exposure model point estimate, correspond well with the exposure estimates as estimated only on the basis of cohort status and their corresponding nitrate/nitrite well water levels. Comparisons of these two sets of data illustrate that following the MHG incident, Cohort II shifted places with Cohort I to become the high-exposure cohort. Further, the predictive ability of the exposure assessment in regard to the outcome of MHG was estimated using a Likelihood Ratio and Pearson's Crosstab analysis. This was performed on the 2-month-of-age point estimate. Likelihood Ratio and Pearson's chi-square were 39.40 and 33.74, respectively, with a probability of achieving these fits by chance alone of <0.0001. This indicates clearly that the children who experienced MHG were also the children at the 2-month-of-age point estimate who had received the highest exposure to nitrate/nitrite through their diet. © 2002 Nature Publishing Group.

Original Publication Date


DOI of published version