Validation of biological markers for quantitative risk assessment.
Authors
Schulte-P; Mazzuckelli-LF
Source
Environ Health Perspect 1991 Jan; 90:239-246
Abstract
The evaluation of biological markers for quantitative risk assessment was discussed. The concept of validity included measurement validity, internal study validity, and external validity. Measurement validity can be defined as the extent to which a measurement means exactly what it claims to measure. Measurement validity can also be characterized as the degree to which a biomarker has content validity, construct validity, and criterion validity. Internal validity was discussed. Internal validity refers to the degree to which inferences drawn from a sample are warranted when the study method, representativeness of the study sample, and the characteristics of the population are considered. Internal validity depends on the ability to control biases, misclassification, and to choose the correct mathematical model for treating the data. External validity was discussed. External validity is the extent to which a study can produce unbiased inferences about the target population and involves the appropriateness of extrapolating between populations or species, from high to low doses, or between different organs in a given species. The concepts of measurement, internal, and external validity were illustrated by applying them to risk assessments of the carcinogenicity of ethylene-dibromide (106934) using bromine (7726956) release as a marker of exposure and the genotoxicity and carcinogenicity of ethylene-oxide (75218) using hemoglobin alkylation as a biomarker. Hemoglobin alkylation was found to be a valid marker for ethylene-oxide exposure. Bromine release was of little value for assessing the risk of ethylene-dibromide. The bromine release data suggested that a threshold exists whereas other data indicated that ethylene-dibromide alkylated macromolecules and caused mutations at doses below the supposed threshold. The authors conclude that before biological markers can be used in risk assessments they must be shown to be reliable. Failure to do so can result in erroneous conclusions.
Keywords
NIOSH-Author; Biological-monitoring; Risk-analysis; Epidemiology; Carcinogenesis; Epoxides; Brominated-hydrocarbons; Alkylating-agents; Hemoproteins
Contact
P. A. Schulte, Industrywide Studies Branch, National Institute for Ocutional Safety and Health, 4676 Columbia Fukway, Cincinnati, OH 45226
CAS No.
106-93-4; 7726-95-6; 75-21-8
Document Type
Journal Article
Source Name
Environmental Health Perspectives
