3.4: At What Noise Threshold Level can Noise Affect Health?
How strong is the research evidence that noise effects follow any particular relationship with the degree of noise exposure? Both field and laboratory data has been used to derive a number of standard dose-effect curves showing very little response at around 40 LDN up to around 100% highly annoyed at around 90 LDN. Figure 2 shows the Schultz curve and its later update relating percent highly annoyed to day-night average sound level (FIDELL, 91). Similar curves have been developed for other effects such as activity and sleep disturbance (see MIEDEMA 98 and PORTER 97). These curves are not summarised in this report. There are of course numerous methodological problems present when attempting to combine the results of a number of different primary studies into an overall meta-analysis and the precise shape of the curves should not therefore be taken too literally.
All curves follow the same generic shape regardless of the particular effect being considered. Due to the scatter of individual data points above and below the curves, the precise shape is often determined more by the method of statistical analysis used than by the actual data. In principle however, noise effects are, or can be assumed to be, low or negligible at low noise levels. At increasing noise levels the effects start to increase slowly at first and then more rapidly. Finally the upper end of the effects scale flattens out at 100%. It is meaningless to consider further increases in the effects of noise above this point where 100% of the population are affected or where individuals are 100% affected. This is shown in figure 3. Whether the shape of the generic S-curve really follows from the available data (given the general statistical uncertainty in this area), or whether it is merely a plausible theoretical construct that happens to fit the data as well (or as badly) as any other theoretical construct is a matter for conjecture.
On the assumption that the lower asymptotes of the curves represent threshold noise exposure values below which effects are infrequent or unlikely, it is possible to derive the guideline values set out in Table 3. For effects other than percent highly annoyed, activity interference and sleep disturbance, the precise shape of the corresponding S-curves is extremely uncertain, even if these hypothetical effects really exist, particularly at the sort of levels typical of environmental or residential exposure. Even for the percentage highly annoyed, it is well known that residents can register strong objections to certain types of noise such as that generated by inconsiderate neighbours even where the noise levels are quite low in absolute terms.
Figure 2: Percentage highly annoyed versus noise level (from FIDELL 91)
Figure 3: Generic curve for noise effects versus noise level
Table 3: Guideline threshold noise exposure values below which reviews report that an effect is unlikely to be observed
Effect Guideline threshold value (dB(A)) Annoyance 40 (LAeq,24h: transportation noise) *1 42 (Ldn: outdoors) *2 55 (dB LAeq: outdoors, few seriously annoyed below this value) speech communication 45-55 (dB LAeq :for elderly or impaired)*4 55-65 (dB Laeq) *4 Sleep 30 (dB Lamax - for continuous noise to avoid serious effects) *4 45 (dB Lamax - for low background and non-continuous noise) *4 waking during the 60 (SEL indoors) *1 night 60 (SEL indoors) *2 60 (dB LAmax )*4 changes to sleep 35 (SEL indoors) *2 stages 40-45 (dB Lamax) *4 subjective reports 40 (LAeq, night) *1 of sleep quality 40 (LAeq, night :outdoors)*2 mood next day 60 (LAeq, night) *1 60 (LAeq, night :outdoors)*2 heart rate 40 (SEL indoors) *2 45 (dB LAmax)*4 Stress related health effects general 65-75 (dB LAeq)*4 cardiovascular effects hypertension 70 (LAeq,, 06-22h :outdoors for road and aircraft traffic noise in living environment) *2 ischaemic heart 70 (LAeq, :outdoors) *1 disease 70 (LAeq,, 06-22h :outdoors for road and aircraft traffic noise in living environment) *2 Hearing loss 70 (LAeq, :indoors for living and recreational environment) *1 65-75 (dB LAeq: "negligible" risk for hearing loss for 8 hour exposure and 40 years age group) *4 Performance 55-65 (dB LAeq: for deteriorated reading acquisition in school children, people learning languages and the elderly) *4 performance by 65 (LAeq, school:outdoors)*1 school children 70 (LAeq, school:outdoors)*2
Key:
*1 (IEH 97), guideline value is the value above which an observable effect might be expected
*2 (NETHERLANDS 97), guideline value is the value of the lowest exposure at which on average an effect has been observed in epidemiological studies
*3 (MORREL 97)
*4 (BERGLUND 96), value is based on (BERGLUND 95) and indicates the value above which an effect is observed[ Previous ] [ Contents ] [ Next ]
Published 12 September 2000
Noise Index
Environmental Protection Index
Defra Home Page
