Noise From Civilian Aircraft in the Vicinity of Airports - Implications for Human Health - Noise, Stress and Cardiovascular Disease

Appendix 1

There are several sets of criteria to assess the likely role of cause in an association observed in an epidemiological study. The criteria used for this report were adapted from the Bradford Hill criteria (Glynn, 1993; Traven, et al., 1995). They are summarized below.

Strength of association. A relative risk greater than 3 provides good evidence that an association is likely to be causal. Important confounders can produce strong associations if not controlled for. A small relative risk, on the other hand, does not rule out a causal relationship - it only means that it is more difficult to exclude some other explanation.

Dose-response. If a dose response relationship is seen, it strengthens the likelihood of causality. Again confounding factors may lead to spurious dose-effect relationships. The absence of a dose-response relationship does not necessarily weaken an association, especially if there is a threshold effect.

Temporality. In its simplest form the causal factor must precede the onset of the disease in question. However, interactions between exposure to the agent and the body's biological system may influence the temporal relationship between exposure and disease.

Reversibility. Removal or reduction of exposure can provide convincing evidence of causality.

Consistency. An association that is shown repeatedly in different studies is unlikely to be due to chance. If conducted by different investigators in different places and times an association is unlikely to be due to a constant bias.

Plausibility. The existence of a suggested mechanism by which the agent causes a disease is reassuring, but will depend on the knowledge at the time. Mechanisms for an association between chronic aircraft noise exposure and cardiovascular disease have been proposed Specificity. Since most diseases have more than one factor contributing to their onset, specificity of an observed association cannot be insisted upon. It is understood that there will not be a one-to-one relationship between aircraft noise and cardiovascular disease.