Contents Search publications                                         Previous  Next

The other way of studying the toxic effects of chemicals is to look at patterns of disease in human populations, such as groups of exposed workers. These are obviously more relevant, because they involve humans rather than animals, but can be difficult to interpret. They can also only be done after human exposure has occurred. But they do allow populations to be followed who have been exposed to chemicals cleared by toxicologists. The main types of epidemiological study are case-control studies and cohort studies.

Case-control studies compare a group of people with a specific disease with another (control) group who do not have the disease but are the same in other respects, like sex, age and social class. Information is then collected on both groups' occupational exposures. Because neither their doctors nor employers are likely to have adequate records of which chemicals workers have been exposed to, and in what quantities, epidemiologists often rely on people's memories. The results are expressed as an odds-ratio (OR). If the OR is significantly greater than 1.0, there could be a link between the exposure and the disease.

Cohort studies look at a group known to be exposed to a particular chemical, and compare their death rate or diseases with a similar, unexposed, group. The results are expressed as Relative Risk (RR). An RR greater than 1.0 suggests a link between the exposure and the disease(s). If a control group cannot be found, comparisons are made between the exposed group and the general population, and expressed as a standardised mortality ratio (SMR).

Epidemiologists weigh the evidence by looking at results from as many studies as possible, as well as the size of the OR RR, the existence of a dose-response relationship, and biological plausibility (is there a believable basis for the exposure leading to the disease?). A scientific consensus may then be reached, as happens with the IARC cancer evaluations.

Contents Search publications                                         Previous  Next