3.5: Linking the Actual Effects of Noise to an Overall Impact on Health
Where reliable exposure response curves exist, and there are of course some doubts on this score in relation to certain effects, we might then assume a direct relationship between the effect of interest and the associated community impact as follows:
exposure -> effect -> impact
The actual situation is rather more complex. Exposure can lead to more than one effect, and community impacts depend on multiple effects as follows:
exposure -> effect ---- | | exposure -> effect -> impact | | exposure -> effect ----There is no agreed method to combine everything into an overall response, even if this were meaningful when taken out of the context of the many and varied social and economic factors that often have much greater health impacts. Most reviewers avoid this problem completely by considering each effect separately, but this is not by any means a satisfactory solution. Recently, Job has set out a preliminary linked model reproduced as Figure 4 (JOB 96). It is debatable whether this really is a comprehensive model or merely a graphical representation of the various inter-relationships that need to be taken into account, but at least it shows that researchers are beginning to tackle some of these essentially quite difficult problems.
Thompson has supported Job's approach by suggesting that effect modifiers could be viewed, not as confounders, but as intervening factors in the causal chain either re-inforcing or diminishing the original stress response (THOMPSON 96). She recommends that future research should tease out those factors which are 'necessary' in the causal chain and which offer evidence that noise exposure precedes the given physiological changes. It is our view that future research which is not specifically addressed to carefully developed and biologically plausible causative hypotheses is unlikely to be able to unravel any of these complex relationships to any useful extent whatsoever. Far too many studies in this general area have merely assumed that some relationship between noise exposure and the health effect under investigation 'exists' and that mere statistical association will be sufficient for the findings of the study to be useful when deciding policy, but there is no evidence that studies of this type can be of any assistance at all.
Table 4 : Vulnerable, susceptible or sensitive groups for each effect (as discussed in the literature)
Effect Susceptible or sensitive groups Annoyance self-reported noise-sensitive individuals*1 noise sensitive people, people with fear of certain sources, those feeling they have no control over the situation have an increased risk of severe annoyance*2 those who indicate fear of aircraft crashes, those concerned with health effects of noise, those who report interference with activities, self-reported noise sensitive individuals*3 general well-being people annoyed by noise in the work place show an increased post-work irritability which might affect their well-being*2 Those with a negative attitude towards noise may be more likely to suffer health effects*5 speech communication elderly and hearing impaired Sleep ill people, older people, people with sleeping difficulties*2 elderly people, shift workers, those with physical or mental disorders people with sleeping difficulties*4 Shift workers*5 waking during the sensitive groups e.g. anxious/depressed*1 night older people Stress related health effects general family history of cardio-vascular disease*5 cardiovascular effects hypertension family history of hypertension*1 people highly annoyed by low levels of road traffic noise*2 people with noise-induced sleep disturbance*2 ischaemic heart men exposed to high levels of traffic noise and disease occupational noise*2 people with noise-induced sleep disturbance*2 Hearing loss effects on the unborn and young children*2 Performance performance by school pupils with learning difficulties, hearing children impairment, English as a second language. *1Key:
*1 (IEH 97)
*2 (NETHERLANDS 97)
*3 (MORREL 97)
*4 (BERGLUND 96),
*5 (JOB 96)Understanding the role of effect modifiers in contributing to outcome variables is important when attempting to assess likely health outcomes after long term exposure to any particular noise environment. Two areas in Job's model relate to reaction modifiers including those relating to individual or group sensitivity (vulnerable or susceptible persons). Various reviewers have suggested ways of identifying vulnerable groups as listed in table 4, but some of the entries reported here do not stand up to close examination. It would be surprising if 'self-reported noise sensitive individuals' did not report higher noise annoyance in noise attitude surveys, but this does not necessarily mean that those same individuals are actually any more significantly affected by the noise. In the absence of any biologically plausible theory as to how noise might actually contribute to hypertension, there is no a priori reason why the class of people with a 'family history of hypertension' should actually be any more sensitive to noise induced hypertension than anyone else. It is equally plausible that people with a family history of hypertension or self-reported noise sensitive individuals might somehow be predisposed towards avoiding higher noise levels that others accept and thereby suffer less health effects as a consequence.
What would be most relevant here would be independently observable variables that might actually modify the strength of the effect being addressed. For example, people with a previous history of ischaemic heart disease or hypertension might be more or less susceptible to future episodes of cardiac ischaemia in the presence of noise than they would otherwise have been. This then becomes a testable hypothesis and is potentially useful in terms of planning future research. To merely ask people how annoyed they are and then relate this to self-reported noise sensitivity is essentially tautologous and not particularly useful for anything other than strictly academic purposes. Other examples might be to compare exposure response relationships between people with high fat and low fat diets, or between people who are measurably more or less reactive to sudden impulsive noise.
Ludlow and Flindell have pointed out that identifying susceptible individuals by means of some independent measure in advance could significantly increase the statistical power of any future research studies in this area (LUDLOW 97). The basic problem which must be overcome for any worthwhile progress to be made is the strong possibility that true effects, if they exist, will be completely swamped by the general statistical uncertainty caused by a combination of the effects being weak, of their affecting only a minority of the population, and of the large number of potential confounding and co-related variables involved. Prior identification of susceptible individuals could go a long way to overcoming some of these limitations. The traditional approach has been to merely consider particular groups as being somehow 'vulnerable' such as old people and young children, even though there is no real evidence in support. It is unlikely that this traditional approach will be at all helpful.
There are further difficulties caused by combinations of noise exposure from more than one noise source. This particularly arises in the case of road traffic noise, which is often present at the same time as other sources. Differences in individual noise exposure attributable to different occupations and life-style are unlikely to be very well represented by measurements of environmental noise taken outside houses. There have been a number of attempts to develop models of combined noise, most recently reviewed by Gjestland (GJESTLAND, 97). There is no consensus at the present time at least in part because no researcher has yet developed a model which can be successfully generalised across to other situations.
Finally, Berglund and others (BERGLUND 95) have suggested that certain essentially mild or trivial noise effects found after short term exposure in laboratory type studies might translate to more significant or serious effects after long term exposure. While this is a plausible suggestion, there is no particular reason to be believe it to be true and there is a great danger of it being used to justify what might otherwise be an unjustifiably precautionary approach. The problem here is perhaps best explained by example. It is well known that short term exposure to relatively high noise levels in laboratory situations can contribute to a number of physiological responses, such as transient elevations in blood pressure, heart rate and certain biochemical indicators. The translation hypothesis then goes on to associate these transient physiological changes with more permanent changes such as hypertension and altered endocrine function, and the case for a precautionary approach is thereby 'proved'. This argument is specious as there are many indications that these same transient physiological responses can be beneficial, particularly when associated with exercise.
In conclusion, it must be recognised that the human organism is extremely complex and it is therefore not surprising that that there are great difficulties in expressing cause-effect relationships in simple terms, particularly when it is remembered that the effects of noise are all indirect in the sense that they represent the way that the body reacts or responds to noise stimuli. There are no direct effects of community and environmental noise on the tissues of the body because the levels of energy involved are far too small.
[ Previous ] [ Contents ] [ Next ]
Published 12 September 2000
Noise Index
Environmental Protection Index
Defra Home Page
