Toxicology vs Epidemiology Toxicology: • Uses lab experiments, typically on animals and with control groups • Role in risk analysis: identify dose-
Toxicological studies occur in a controlled laboratory Epidemiological studies usually involve a comparison between two groups of individuals
At first sight the question may seem to be a rhetorical one - we are so accustomed to look upon these two disciplines as complimentary for the assessment of
18 juil 2011 · Toxicology and Epidemiology: Improving the Science with a Framework example, geographical differences or changes in disease
UNIT III--INTRODUCTION TO TOXICOLOGY AND RISK ASSESSMENT Implicit in the practice of epidemiology is the need for the different dis-
Toxicant, toxin, and poison are often used interchangeably in the literature; however, there are subtle differences as indicated in the table Toxic substances
![[PDF] Toxicology vs Epidemiology](https://pdfprof.com/EN_PDFV2/Docs/PDF_7/76823_7ToxvsEpidem.pdf.jpg "[PDF] Toxicology vs Epidemiology")
76823_7ToxvsEpidem.pdf
Toxicology vs Epidemiology
Toxicology:
•Uses lab experiments, typically on animals and with control groups. •Role in risk analysis: identify dose-response relationship.
Epidemiology:
•Usually gathers and analyzes data from the real world.
Problems in Toxicological
Modeling
•High to low dose extrapolation •Species-to-species comparisons •Exposure route comparisons (stomach, lungs, skin). •Interactions among multiple toxins
High to Low Dose Extrapolation:
Carcinogenesis
•High dose effects may produce carcinogenic mechanisms not found at low doses ~> overestimation of risk. •Early, low-dose tumors are not identified as such ~> underestimate of risk. •Tumors might have appeared after sacrifice ~> underestimate of risk •Cumulative and sporadic doses may have different effects. •Model selection may produce over or under estimation of risk.
Model Choice Affects Low to
High Dose Extrapolation
•Threshold or not? •Two-hit model? (production of oncogenes plus disabling of tumor suppressor gene - order of mutations not important) •Multistage? e.g., initiation (reversible; oncogene or tumor suppressor disabling), promotion (proliferation), progression (spread to other locations). •Different models may fit different types of cancer.
Species to Species Comparison
•Different body weights have different metabolisms, which may affect processing of toxins. •Mechanisms and target organs or tissues may not be analogous. •Use of homogeneous and highly susceptible test animals is necessary to obtain a detectable response and produce health-conservative risk estimates, but this may inflate cancer rates.
Reasons for High-Dose
Toxicological Assessments
•Modeling provides information about possibilities. (e.g., if cost estimates for remediation vary by several orders of magnitude, but all are low, the choice of action is clear.) •Modeling provides information about which assumptions are most important and highlights where more research can be most useful. •Can address causality by using control groups. •Experiments with small doses need large numbers of animals, which is prohibitively expensive. (High dose studies may cost $1 - 5 million.) •Experiments with humans are unethical.
Problems with Epidemiological
Studies
1.Can not address cause and effect
•No control group •Confounding variables (appropriate analysis helps somewhat)
2. Measurement error
•Self-report may be necessary •Estimates, averages, or surrogates may be necessary.
Problems with Epidemiological
Studies, Continued
3. Extrapolation
•Population studied may have higher exposure than general population •Population studied may not be representative of general population (health, age, etc.)
4. Interpreting small effects (e.g., RR < 2)