Field studies and laboratory experiments on noise-induced sleep disturbance show conflicting results and do not provide sufficient knowledge for valid exposure-effect relationships. It is also well known that habituation exists for awakenings whereas other effects such as heart rate reactions and minor EEG-reactions do not habituate. In this paper comparisons are made between findings from a series of laboratory and field studies on effects of road traffic noise on sleep performed by this athor. Possible reasons for the discrepancies found in various studies in laboratory and field are commented on and methods and research needs are discussed. Comparison of results obtained in research on perceived sleep quality parameters at this Department that used the same methods in laboratory and field studies showed fairly good agreement for difficulties in falling asleep and perceived sleep quality whereas awakening reactions were much less frequently reported in the field studies.
Keywords: perceived sleep quality, awakenings, body movements, road traffic noise, field, laboratory
|How to cite this article:|
Ohrstrom E. Sleep disturbances caused by road traffic noise - studies in laboratory and field. Noise Health 2000;2:71-8
| Introduction|| |
Extensive reviews on international sleep and noise research have been presented in recent years, see e.g. papers presented at the ICBEN Conference in Sydney, 1998. This paper focuses on sleep disturbance and road traffic noise and gives an overview of this author's own research on sleep in laboratory experiments and field investigations since the early 1980s. The paper concentrates on perceived sleep quality parameters and behavioural awakenings. Habituation, dose-response relationships and possible reasons for discrepancies between results from laboratory experiments and field studies are discussed. For detailed results see articles in the reference list.
Results have been obtained from 6 series of laboratory experiments on 734 subject nights involving 82 young (18 - 34 years of age, on average 24.3), healthy subjects of both sexes.
The aim of early laboratory studies (Ohrstrom and Rylander, 1982) was to develop objective and less expensive methods that could be used both in laboratory and field studies without interfering with subjects' sleep. We chose to study body movements with an accelerometer fastened to the bed, perceived sleep quality parameters, mood and performance.
Two series of experiments were executed with six and 12 subjects respectively. The subjects in experiment 1 were exposed to intermittent noise (37 noise events with a maximum noise level of 80 dBA) and steady, non-fluctuating road traffic noises at the same LAeq level (51.4 dBA). Experiment 2 used intermittent road traffic noise at two different LAmax-levels (60 and 70 dBA). The main findings from these experiments were that intermittent noise caused significantly more sleep disturbance than steady, non-fluctuating noise. Sleep quality decreased with increased LAmax levels. [Table - 1] shows the results on perceived sleep quality and perceived number of awakenings.
A redistribution of body movements was found during nights with intermittent noise as compared with quiet reference nights. The number of movements increased during the time periods with the highest number of noise events (r=0.86 between movements and the hourly LAeq level).
To study habituation effects and the importance of subjective noise sensitivity for sleep disturbance effects, a two-week laboratory experiment (Ohrstrom and Bjorkman, 1988) was performed with 24 subjects. They were exposed to 37 noise events with an LAmax level of 60 dBA during eight of 13 consecutive nights. The main results on perceived sleep quality in experiment 3 are shown in [Table - 2].
A significant habituation effect was seen for awakenings (p<0.001) among the sensitive group. For non-sensitive persons, there was an increase in tiredness (p<0.05) after the last exposure night.
In these experiments no habituation was found for either of the two groups in physiological measures on heart rate (beats/minute) or number of body movements.
Three other series of experiments (experiment 4, 5 and 6) focused on the importance of the number of noise events at different LAmax levels (Ohrstrom and Rylander, 1990 and Ohrstrom 1995). In two experiments (experiment 4 and 5), 28 subjects were exposed to intermittent road traffic of either 50 or 60 LAmax from 4, 8, 16 and 64 events during four of eight consecutive nights. In one experiment (experiment 6), 12 other subjects (rather or very sensitive) were exposed to 16, 32, 64 and 128 noise events with a LAmax level of 45 dB.
The results on perceived sleep quality parameters for the three experiments are shown in [Figure - 1] a-c. All test sessions revealed increased difficulties in falling asleep with an increased number of events. Difficulties in falling asleep seemed to be more closely related to number of events than to maximum noise level [Figure - 1]a. The number of awakenings (mean average) was related to LAmax levels and to number of noise events, but a significant increase compared with reference nights was present only for 60 LAmax at 4,16 and 32 noise events. The percentages of subjects who reported awakenings due to noise, increased with number of noise events even at the lowest maximum noise level (45 LAmax) [Figure - 1]b. As for perceived sleep quality, a significant decrease was found from 32 events for 45 LAmax (- 9%) and from 16 events for 60 LAmax (- 29%) [Figure - 1]c.
In the experiments 4, 5 and 6, a tendency toward a decrease was found for body movement reactions in connection with noise events at higher numbers of noise events. This is shown in [Figure - 2].
The decrease in body movement reactions at 64 noise events for 50 and 60 LAmax as compared with nights with less number of noise events is statistically significant (p<0.03 Mann-Witney U-test).
Results have been obtained, by this author, in four field studies involving 781 persons of both sexes, aged 18-75 years. The main aim of these studies was to study various adverse effects of road traffic and not primarily sleep disturbances effects.
To study long-term effects of noise, a crosssectional study was designed to elucidate various adverse effects of noise in terms of annoyance, sleep disturbance and possible long term consequences for general wellbeing (Ohrstrom, 1991). Significantly greater difficulties in falling asleep, greater tiredness in the morning, lower physiological and mental wellbeing and a tendency toward decreased sleep quality was found in a noisy area (72 LAeq 24 h) compared with a control area (<50 LAeq).
Another study was designed to study effect exposure relationships in which six areas with different numbers of heavy vehicles during night hours were selected for investigation (Ohrstrom, 1993). No relation was found between sleep quality parameters and outdoor noise level in LAmax, number of noise events or LAeq level. When the sample was divided into two groups (bedroom window facing/not facing road), significant differences in sleep quality were found.
In a field study (Ohrstrom, 1989) carried out at different distances from a highly trafficked road (noisy area LAeq 24 h 72 ) and a control area (LAeq <55), significant differences were observed for perceived sleep quality, difficulties to fall asleep, awakenings and tiredness early morning and day. Various physiological and mental symptoms (headaches, decreased social orientation, nervous stomach and tiredness) were reported significantly more frequently in the noisy area. The study was repeated 11 years later, in 1997 (Ohrstrom and Bjorkman,1998), with similar results. In this study noise measurements were made in a number of bedrooms with different types of windows. The measurements showed that the facade reduction varied between 29 and 38 dB depending of type of window.
Comparisons between laboratory and field studies
Despite minor differences in methods and uncertainties about the indoor noise levels in the field studies, results from these laboratory and field studies can be compared for the parameters; difficulties in falling asleep, awakenings and perceived sleep quality. Results on perceived sleep quality are shown in [Figure - 3]. Perceived sleep quality is expressed in terms of changes in percentage from reference nights (laboratory studies) and from control sites (field studies).
The figure shows that sleep quality as measured in laboratory settings decreases with noise level and that the decrease is larger at 64 events (lower curve) than at 16 events (upper curve). Three of the data points from the five field studies indicate larger decrease in sleep quality than the 16 events curve from the laboratory settings but smaller decrease than the 64 events curve. One field data point (G91) shows less decrease in sleep quality compared with results from laboratory settings and data point (B98) falls just above the 16 events curve.
When results from laboratory and field studies were compared for difficulties to fall asleep all field data points fell on the 16 events curve. For awakening reactions much lower effects were however found in the field studies than in the laboratory (all field data points below the 16 events curve).
There are large uncertainties regarding the noise level of the field data and they should thus be treated with care, because indoor noise levels, LAmax and number of events are extrapolated (except for point B98) from outdoor measurements in representative positions in each study area (not individual measurements). The position of the bedroom window was controlled for in only one of the field studies. In these extrapolations, facade reduction was estimated to 25 dB and the difference between road and yard was estimated to 10 dB.
It can be concluded from the studies reviewed here that awakening reactions are much fewer in field studies than in laboratory settings whereas there are smaller differences for perceived sleep quality. In their review, Pearsons et al., 1995 compared results on sleep disturbance effects (awakenings and sleep stage shifts) from four field studies and 15 laboratory or quasilaboratory studies and they also found much greater effects in laboratory studies than in the field, especially for awakenings. There are, of course, several reasons for these discrepancies. One possible reason is different sources of noise exposure in the studies reviewed by Pearsons et al. Of the 15 laboratory studies only two used road traffic noise (others concerned sonic boom, tones and jet flyover etc) and two of the four field studies concerned road traffic noise. Another important reason for the discrepancies between findings in laboratory and field is probably the lack of control (or knowledge) of individual indoor noise levels in many field studies. This is a severe problem when studying effect-exposure relationships as opposed to laboratory experiments where the same persons are exposed to carefully controlled noise exposure situations. If the indoor noise levels in field studies are unknown, comparisons between field and laboratory results on sleep disturbance effects are therefore highly uncertain.
In the studies reviewed here by the present author, perceived sleep quality as measured in field studies deviated much less from laboratory results than awakenings. If the noise levels in the field studies were overestimated (which is more likely than the opposite in the present studies) by 5-10 dBA, e.g. 30 dB facade reduction instead of 25 dB or bedroom windows facing the yard instead of road, the discrepancy between these laboratory and field studies in perceived sleep quality would be very small. The awakenings would however still be fewer in these field studies. It is well known that habituation exists for awakenings whereas other effects such as heart rate reactions and minor EEG-reactions do not habituate. Thus, another possibility for the large discrepancies found by Pearsons et al is, which is in accordance with Thiessen's 1978 findings (and Ohrstrom 1988), that habituation occurs for awakenings but not for sleep stage shifts and perceived sleep quality. Thiessen found in his 24-night laboratory study on road traffic noise that the probability of (behavioural) waking decreased to half the value in about two weeks, whereas no habituation was seen for shifts to a more shallow sleep stage.
Field studies and laboratory experiments on noise-induced sleep disturbance show more or less conflicting results depending on type of sleep disturbance criteria and do not provide sufficient knowledge for valid exposure-effect relationships. For the assessment of accurate exposure-effect relations, more extensive and reliable studies must be conducted with much better control, or knowledge, of indoor noise immission and type of noise. Although longterm effects of noise-induced sleep disturbances must be studied in the field, preferably in longitudinal studies, results from laboratory experiments are valuable, not only for the development of methods but also for studying acute exposure-effect relations under carefully controlled exposure conditions. Exposure-effect relationships also need to be determined in a comparative way for aircraft, road-traffic and railway noise and noise indices suited for sleep disturbance must be selected. In order to establish threshold levels and exposure-effect relationships, good indicators of sleep disturbance must also be determined. Annoyance is generally accepted as an effect criterion for the assessment of guidelines for daytime noise exposure. In accordance with this, both objective and subjective evaluations of sleep quality ought to be taken into account as a basis for the assessment of guidelines for night time noise. Potential indicators are time taken to fall asleep, effects of noise during sleep, premature awakening, perceived sleep quality, and aftereffects of noise induced sleep such as sleepiness, mood and performance after exposure to nighttime noise. The relation between "objective" or physiological measures and perceived sleep is as yet unclear. It is, however, well documented (see e.g. Middelkoop et al., 1996) that there are significant correlation between various perceived sleep measures. These authors demonstrated that self-estimated sleep latency time is the most important determinant of perceived sleep quality, followed by nocturnal awakenings and shorter total sleep time (together these measures accounted for 45 % of the variation in sleep quality).
Person-linked factors are important for sleep and may affect exposure-effect relationships. The proportion of a population that is susceptible to noise-induced sleep disturbance must be determined. Information has also to be collected on differences in sleep behaviours and sleep disturbance due to geographical and cultural factor characteristics, factors that may influence exposure-effect relationships.
| References|| |
|1.||Middelkoop, A M et al. (1996) Subjective sleep characteristics of 1484 males and females aged 50-93: Effects of sex and age, and factors related to self-evaluated quality of sleep. Journal of Gerontology: Medical Sciences, vol 51A, (3),M108-M115. |
|2.||Pearsons et al. (1995) Predicting noise-induced sleep disturbance JASA 97(1) p 331-338 |
|3.||Thiessen G. (1978) Disturbance of sleep by noise. JASA, 64, pp 216 -222. |
|4.||Ohrstrom E. and Rylander R. (1982) Sleep disturbance effects of traffic noise - a laboratory study on after effects. J Sound Vib 84, 87-103. |
|5.||Ohrstrom E. and Bjorkman M. (1988) Effects of noisedisturbed sleep - a laboratory study on habituation and subjective noise sensitivity. J Sound Vib 122 (2), 277-290. |
|6.||Ohrstrom E. and Rylander R. (1990) Sleep disturbance by road traffic noise - a laboratory study on number of noise events. J Sound Vib 143, 93-101. |
|7.||Ohrstrom E. (1991) Psychosocial effects of traffic noise exposure. J Sound Vib 151 (3) 513-517. |
|8.||Ohrstrom E. (1995) Effects of low levels from road traffic noise during night - a laboratory study on number of noise events, maximum noise levels and noise sensitivity. J Sound Vib 179, pp 603-615. |
|9.||Ohrstrom E. (1989) Sleep disturbance, psychosocial and medical symptoms - a pilot survey among persons exposed to high levels of road traffic noise. J Sound Vib 133 (1), 117-128 . |
|10.||Ohrstrom E. (1993) Long-term effects in terms of psychosocial wellbeing, annoyance and sleep disturbance in areas exposed to high levels of road traffic noise. In M Vallet (Ed.) Noise and Man '93, Noise as a Public Health Problem, Proceedings of the 6th International Congress. Institut National de Recherche sur les Transports et leur Securite Nice, France 5-9 Juillet 1993. Volume 2, 209-212. |
|11.||Ohrstrom E. and Bjorkman M. (1998) Sleep disturbances before and after reduction of road traffic noise. In Noise Effects ´98. Proceedings of the 7th International Congress on Noise as a Public Health Problem Sydney Australia 2226 November 1998, Volume 2, pp 451 - 454. |
Department of Environmental Medicine, Göteborg University, Göteborg
Source of Support: None, Conflict of Interest: None
[Figure - 1], [Figure - 2], [Figure - 3]
[Table - 1], [Table - 2]