Skip directly to search Skip directly to A to Z list Skip directly to page options Skip directly to site content

NIOSHTIC-2 Publications Search

Search Results

Your results may vary: exploring the sensitivity of titanium dioxide risk estimates to different modeling assumptions.

Dankovic-DA; Kuempel-ED; Wheeler-MW
Toxicologist 2009 Mar; 108(1):305
Risk analysts typically make a number of decisions during the conduct of a quantitative risk assessment, and these decisions may have significant impact on the final risk estimate. In this analysis we explore the sensitivity of risk estimates for titanium dioxide (TiO2) to pooling of data, dose-response model selection, animal-tohuman extrapolation method, and the choice of a human dosimetry model. The goal was to estimate the occupational exposure concentration of TiO2 associated with a 0.1% lifetime excess risk of lung tumors, based on long-term bioassay data for both fine (pigment-grade) and ultrafine (UF) TiO2. The method used was benchmark dose (BMD) modeling, estimating either the model-based 0.1% exposure concentration in rats, or a 10% "point of departure" coupled with linear extrapolation. Potential modeling decisions included use of the best-fitting model, the most health-protective model, or a model average, and whether to use the central BMD estimate or a 95% lower-bound (BMDL). Extrapolation issues include the choice of dose metric -- either particle surface area/g-lung or particle surface area/lung surface area -- and the choice of a human lung particle dosimetry model. The BMD estimates vary by a factor of 29 between the most health-protective dose-response model and the best-fitting, while the BMDLs differ by 2.7. The BMD and BMDL differ by factors of 1.3-14, depending on the choice of dose-response model. Pooling fine and UF TiO2 data has little impact on the risk estimates for fine, but does affect the risk estimates for UF. The choice of dose metric for ratto- human extrapolation impacts the risk estimate by roughly 3-fold, and the choice of dosimetry model alters it by approximately 2-fold. Overall, the estimated occupational exposure associated with a 0.1% excess risk of lung tumors may vary more than 100-fold, depending on the specific assumptions which are made. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health.
Biological-factors; Dose-response; Dosimetry; Exposure-assessment; Exposure-levels; Inhalation-studies; Laboratory-animals; Lung-function; Lung-irritants; Mathematical-models; Quantitative-analysis; Respiratory-hypersensitivity; Respiratory-system-disorders; Risk-factors; Statistical-analysis; Nanotechnology
Publication Date
Document Type
Fiscal Year
NTIS Accession No.
NTIS Price
Issue of Publication
NIOSH Division
Priority Area
Source Name
The Toxicologist. Society of Toxicology 48th Annual Meeting and ToxExpo, March 15-19, 2009, Baltimore, Maryland