Standard Article

Essentials of Epidemiology for Toxicologists

Specialisation

  1. David Coggon OBE, MA, PhD, DM, FRCP, FFOM, FFPH, FMedSci Professor of Occupational and Environmental Medicine

Published Online: 15 DEC 2009

DOI: 10.1002/9780470744307.gat111

General, Applied and Systems Toxicology

General, Applied and Systems Toxicology

How to Cite

Coggon, D. 2009. Essentials of Epidemiology for Toxicologists. General, Applied and Systems Toxicology. .

Author Information

  1. University of Southampton, MRC Epidemiology Resource Centre, Southampton, UK

Publication History

  1. Published Online: 15 DEC 2009

Abstract

Epidemiology is concerned with the distribution and determinants of health in populations. The applications of epidemiology that are of most relevance to toxicology lie in the identification and characterization of toxic hazards, the assessment of risks from toxic exposures, and the evaluation of measures designed to control such risks. Most epidemiological studies focus on the occurrence of disease, but other health outcomes such as symptoms and disability may also be investigated.

Incidence (the rate at which new cases of disease occur in a population) is the measure of disease frequency of most relevance to study of causation. Other measures such as mortality (the incidence of death from a disease) and prevalence (the proportion of a population who are cases at a point in time or during a specified period) may be used as a proxy for incidence, but results must then be interpreted with added care. Where disease rates vary importantly by sex and age, comparison between populations may be enhanced by use of sex- and age-specific rates, or by standardizing rates for sex and age.

Various statistics are used to summarize associations between ‘risk factors’ and health outcomes, each with its particular applications. Attributable risk is relevant when making decisions in risk management for individuals. Relative risk (RR) (or the closely related odds ratio) is most useful when considering whether associations are likely to be causal. Population attributable risk and attributable proportion are useful in risk management for populations. The attributable fraction in exposed (AFexp) is used when determining causal attribution for purposes of compensation.

Major considerations in the design and interpretation of epidemiological studies are bias (a systematic tendency to underestimate or overestimate a parameter of interest), chance and confounding (which occurs when the risk factor of interest is associated with a ‘confounding’ factor that independently determines the risk of developing the health outcome under study). Assessment of the potential impact of chance is helped by statistical inference using either hypothesis-testing (p-values) or confidence intervals.

Categories of epidemiological investigation include descriptive studies, ecological studies, cohort studies, case–control studies, cross-sectional surveys and randomized and nonrandomized experiments. Each of these study methods has its particular applications, strengths and limitations.

In addition to bias, chance and confounding, other considerations when interpreting and comparing epidemiological findings are the ways in which exposures and health outcomes have been classified, and the potential for biological modification of causal associations by ‘effect modifiers’. In addition, epidemiological results should always be viewed in the context of relevant biology, including what is known from toxicological studies in the laboratory.

Keywords:

  • epidemiology;
  • design;
  • interpretation;
  • applications;
  • risk assessment