Risk Assessment, Epidemiology, and Regulatory Practice
Four steps and their numbers
Risk assessment comprises hazard identification, dose-response assessment, exposure assessment, and risk characterization. For noncancer endpoints, the reference dose (RfD) equals a point of departure (NOAEL or benchmark dose lower confidence limit) divided by uncertainty factors, typically 10 for animal-to-human, 10 for sensitive subpopulations, and 10 for database gaps, with modifying factors added when warranted. For cancer, a slope factor supports linear low-dose extrapolation, and risk = dose × slope factor. Margin of exposure = point of departure/exposure; 100 or more is often considered acceptable. Epidemiology supplies evidence through cohort, case-control, cross-sectional, and ecological designs, using relative risk, odds ratio, attributable risk, bias, confounding, and effect modification, and judged with Bradford Hill considerations. Regulation involves EPA, FDA, OSHA, IARC, NTP, GHS, REACH, and TSCA.
Traps: the RfD addresses noncancer effects and the slope factor addresses cancer; mixing them is a classic error. A NOAEL is not a threshold, and the RfD adds uncertainty factors. Statistical significance in an epidemiological study is not causation. IARC classifications describe strength of evidence, not potency.
Rule of thumb: RfD = NOAEL/UFs. If NOAEL = 10 mg/kg-day and UFs are 10 × 10 = 100, RfD = 0.1 mg/kg-day.