A descriptive cross-sectional study of annoyance from low frequency noise installations in an urban environment :<b>K Persson Waye, J Bengtsson, A Agge, M Bjorkman</b>, Noise Health

Home

ARTICLES	[Download PDF]

Year : 2003 \| Volume : 5 \| Issue : 20 \| Page : 35--46

A descriptive cross-sectional study of annoyance from low frequency noise installations in an urban environment

K Persson Waye, J Bengtsson, A Agge, M Bjorkman
Department of Environmental Medicine Göteborg University, Göteborg, Sweden

Correspondence Address:
K Persson Waye
Department of Environmental Medicine, Box 414, 405 30 Göteborg
Sweden

Abstract

In order to improve the living conditions for respondents highly exposed to traffic noise, it has been recommended that one side of the building should face a «DQ»quiet side«DQ». Quiet may, however, be spoilt by noise from installations such as ventilation and air-conditioning systems. The noises generated by installations of this kind often have a dominant portion of low frequencies (20-200 Hz) and may be a source of great annoyance and sleep disturbance. This paper describes the cross-sectional part of an intended intervention study among residents exposed to traffic noise on one side of the building and to low frequency noise from installations on the other side of the building. A questionnaire masked as a general living environment study was delivered to a randomly selected person in each household. In total 41 respondents answered the questionnaire (71% response rate). Noise from installations was measured indoors in a bedroom facing the courtyard in a selection of apartments and outdoors in the yard. 24h traffic noise outdoor and indoor levels were calculated. The noise levels from installations were slightly above or at the Swedish recommendations for low frequency noise indoors with the window closed and exceeded the recommendations by about 10 dB SPL when the window was slightly opened. The proportion of persons who reported that they were very or extremely annoyed indoors from noise from installations was more than twice as high as for traffic noise. Installation noise also affected respondents' willingness to have their windows open and to sleep with an open window. The high disturbance of installation noises found in this study indicates the importance of also regulating the noise exposure on the «DQ»quiet side«DQ» of buildings. Further studies will give a better base for the extent of annoyance and acceptable levels of installation noises.

How to cite this article:
Persson Waye K, Bengtsson J, Agge A, Bjorkman M. A descriptive cross-sectional study of annoyance from low frequency noise installations in an urban environment.Noise Health 2003;5:35-46

How to cite this URL:
Persson Waye K, Bengtsson J, Agge A, Bjorkman M. A descriptive cross-sectional study of annoyance from low frequency noise installations in an urban environment. Noise Health [serial online] 2003 [cited 2023 Mar 26 ];5:35-46
Available from: https://www.noiseandhealth.org/text.asp?2003/5/20/35/31690

Full Text

Background

To minimize the adverse effects of traffic noise on health and sleep among residents exposed to high traffic noise levels, it has been recommended that one side of the building should be a "quiet side" (Kihlman, 1993]. However, the quietness may be spoiled by noise from installations such as ventilation, heating and air-conditioning systems, which are often positioned on the side of the building not facing the street. Noises from such installations often have a dominant portion of low frequencies (20-200 Hz) and may be a source of great annoyance and sleep disturbance (see Berglund et al., 1996, for a review). The adverse effects of low frequency noise have been reported in a large number of case studies (e.g. Bryan, 1976; Challis and Challis, 1978; Chatterton, 1979; Cocchi and Fausti et al., 1992). A limited number of epidemiological studies also give some support to the findings of the case studies. Verzini and Frassoni et al. (1999) found that the energy content of the frequency band 20 to 160 Hz was significantly related to sleep disturbance, concentration difficulties, irritability, anxiety and tiredness. The study was carried out among 98 subjects living in urban areas with dominant low frequency noise from installations, air condition units, industrial processes and traffic noise from tunnels. Similar symptoms, except for anxiety, were related to annoyance from low frequency noise in a Swedish cross sectional study comprising a total of 279 people exposed to low or flat frequency noise from ventilation/heat pumps in their homes (Persson Waye and Rylander, 2001). The prevalence of annoyance by low frequency noise was statistically higher and ranged from 15 to 20% as compared to a prevalence of 0 to 4% in reference areas, with ventilation noise of a flat frequency character. The significant difference between the exposed group and the reference group remained when correction was made for differences in Aweighted noise levels. In another investigation, 30 subjects complaining of low frequency noise in their homes were compared to an equal number of subjects of matched age and sex, living in the same block of flats but without the low frequency noise (Mirowska, 1998). A higher occurrence of chronic sleep disturbance and depression was reported among the complainers. The study gives some indications of higher symptoms among complainers, but the results could be confounded by differences between the study populations.

The studies referred to above were carried out with some methodological differences, the most important one related to the assessments of the noise exposure. Of the studies, only two carried out recordings indoors (Persson Waye and Rylander, 2001; Mirowska, 1998). For low frequency noise, it is highly important to relate effects to measurements made indoors, as the attenuation of the facade and the room dimensions will affect the resulting noise indoors. In spite of the methodological differences the investigations do show that low frequency noise in the living environment causes annoyance, while the effects of long-term exposure to sleep and health are less well explored.

This paper describes the first part of a study carried out among residents exposed to traffic noise on one side of the building and to low frequency noise from installations on the other side of the building. The purpose was to evaluate the prevalence of annoyance at home caused by low frequency noise emitted from installation noises outdoors. In subsequent phases of the project, effects on annoyance, health and sleep disturbance will be studied among similar populations in the city of Goteborg and compared to referent groups not exposed to low frequency installation noise. Studies will also be undertaken among some of these populations after actions have been undertaken to reduce installation noises, in order to investigate effects on annoyance, health and sleep disturbance.

Methods

General outline

A cross sectional study was carried out among tenants in blocks of flats with one side facing a trafficked street and the other side facing a courtyard with a large number of fans, compressors and air cooling systems. Subjective responses were collected by questionnaires. Noise levels from the installation sources were measured indoors and outdoors and noise levels from traffic were calculated.

Study area

The study area was three buildings comprising blocks of flats, A, B and C, surrounding a courtyard on three sides [Figure 1]. A fourth building, which faced the courtyard, was an office block and was not included in the study. The flats faced the courtyard on one side and a trafficked street on the other. The traffic flow on the street outside block A was 1000 vehicles per 24 h with 5% heavy vehicles, and the traffic flow on the streets outside blocks B and C was 14 500 and 16 500 vehicles per 24 h, respectively, with 4-5% heavy vehicles. The 24h period was an average for a period calculated with data from one calendar year.

In the courtyard, 35 different installations were located on the ground or on the roof of the buildings. The fans and compressors were extracting air or cooling air from restaurants and small enterprises housed in the ground floor of the buildings. The noises generated were of a predominantly low frequency character.

Source population

The source population comprised a randomly selected person in each household between the ages of 18 and 80 years who had lived in the apartment for a minimum of one year. There were 69 apartments in the three buildings. Six respondents were excluded from the source population because of age, because they did not speak Swedish or because they were hospitalised. Six apartments were not rented on a permanent basis and were therefore also excluded. The total source population thus included 57 households.

Questionnaire

A questionnaire masked as a general living environment study was delivered to the door together with an introduction letter describing the purpose of the study. The respondents were asked to post the questionnaire but, as the response rate after two reminders was still low, we tried to contact them in a personal visit or by telephone. Forty-one questionnaires were returned, which gave a response rate of 71%.

The questionnaire had 35 questions divided into general questions on satisfaction with the dwelling standard and the living environment, disturbance from different environmental factors such as exhaust from motor traffic, cooking smells and noise from music from nearby restaurants, noise from fans and compressors and traffic noise. The questions on noise annoyance from traffic and fans/compressors were posed as: "Thinking about the last three months when you are at home, how much does noise from…….annoy you?" The question was a Swedish translation of the question proposed by Fields et al., (2001). The same phrase was used to ask about annoyance outdoors but with reference to when the respondent spent time outdoors in the courtyard. Questions were also posed on disturbance of rest and relaxation. The answers were given on a five-grade verbal scale, with the alternatives not at all, slightly, moderately, very and extremely. The questions on annoyance were answered on an opinion scale from 0 to 10, where 0 was equivalent to not at all annoyed and 10 equivalent to extremely annoyed (Fields et al., 2001). A section of the questionnaire comprised questions on sleep quality, sleep behaviour and sleep disturbance followed by questions on health. Questions were also posed on how long the respondents had lived at their present address, their occupation, work hours, the number of members in their family, age and sex. The study was carried out in August and September 2000.

In this presentation, descriptive results will be given of the prevalence of effects related to noise disturbance and sleep disturbance. Symptoms that may also be strongly related to other socioeconomic factors are not reported, as the sample is too small for conclusions to be drawn about those effects and as there is yet no reference group with which comparisons can be made.

Assessment of noise exposure

Noise measurements were made after the questionnaire study was completed. The low frequency sound pressure levels were measured indoors in a bedroom facing the courtyard. Measurements were made in ten different apartments spread equally over the three blocks and representing high and low floors within the range of the 1st to the 5th floor. Measurements were recorded with the window closed and with the window slightly open (5 cm). The measurements were carried out at three positions in the room, one of them where the highest Cweighted noise level was found, in accordance with (SP-REPORT 1996:10). The sounds were recorded for two minutes at each position using a real-time analyser (B&K 2260). The sounds were also stored on digital tapes on a DAT recorder (SONY TCD-DC7). Subsequent analysis of equivalent third octave band sound pressure levels was done within the frequency range of 20 to 10 000 Hz. In the analysis, the three measurement positions in each apartment were in accordance with SP REPORT (1996:10), logarithmically averaged. The different apartments' average levels were then arithmetically averaged, and standard deviations thus represent deviations between flats rather than deviations inside the room. The reason for choosing an arithmetic average of the different flats sound pressure levels was to obtain a measure that would be representative for the area.

The noise levels from the installations were recorded outdoors in the courtyard and on one balcony on the fifth floor over a period of two months (September and October). The equivalent A- and C-weighted levels and the statistical distribution during one-hour intervals were recorded. A remote controlled measuring system (Larsson and Davis model 820) was used, and the microphone B&K 4165 was placed at a height of two meters. The recordings on the balcony were made with the same system, but the microphone was positioned one meter in front of the facade. These recordings were corrected for a facade reflex of 3 dB to allow comparisons with the free field measurements.

Equivalent 24h traffic noise outdoor levels were obtained from calculations made at the local health and environmental authorities. The calculations were made in accordance with the Nordic calculation model (Jonasson and Nielsen, 1996). Calculated levels were obtained at each floor and also included estimations of indoor noise levels based on a 27 dBA reduction by the facade. Noise measurements were also undertaken on two balconies facing the streets during a period of a week, in order to obtain the distribution of traffic noise levels during daytime, 07.00-19.00, evening 19.00-23.00 and night time. On the basis of the distributions obtained, and the 24h calculated noise levels, separate noise levels were calculated for daytime, evening and night time.

Statistical treatment of data

The prevalence of reports of disturbance was analysed for the total sample and for the sample subdivided into respondents with bedrooms facing the courtyard (CY) and respondents with bedrooms and sitting rooms facing a street (ST). 95% confidence intervals of proportions were calculated according to Altman (1991). Differences between subgroups were tested with the Chi-square or Mann Whitney U-test. Relationships between variables were tested using Spearman's correlation analysis. All tests were two sided and a p-value below 0.05 was considered statistically significant.

Results

General data

The study population comprised 25 households (59%) with one person and of 16 households (39%) with two or more persons. Five families had children below the age of 18. The respondents had lived at their present address between 1 and 59 years, the median value being four years. The majority of the respondents (63%) had lived in urban areas before they moved to the present address. Most respondents (88%) were very satisfied or satisfied with their apartments and 71% were very satisfied or satisfied with their living environment. Nearly all respondents, or 98%, answered, however, that they seldom or never used the courtyard for relaxation purposes. Living conditions were considered to be good by 68%, while 29% regarded conditions as not particularly good or bad.

Seventy-one percent of the respondents reported that alterations ought to be made in their home and living environment. The majority of the comments were related to improved conditions in the flats (n=9), less traffic (n=6), reductions of the noise from fans/compressors (n=4) and better handling of garbage (n=4).

Of the 41 respondents, 19 had their bedrooms facing the courtyard, 20 had bedrooms and living rooms facing a street and two had bedrooms facing both the courtyard and a street. These two latter respondents were not classified into either of the categories.

The median age was 43 years in CY and 30 years in ST; the difference was not significant (z=1.695, p= 0.091). In the CY sample the proportions of men and women were similar, 47% vs 53%, while the ST sample comprised a somewhat higher proportion of women (75%). This difference was however not significant (=2.119, p=0.146). The distribution of other socio-economic factors did not differ between the respondents in the two subgroups.

Among the respondents with bedrooms facing the street, six lived in block A and 12 lived in block B or C and two had bedrooms facing both streets.

Noise exposure

Blocks B and C faced a heavily trafficked street on one side, while block A faced a relatively moderately trafficked street. The levels of installation noise were comparable for blocks B and C while respondents living in block A were exposed to a somewhat lower level. The data on noise measurements were thus calculated for blocks B and C together and for block A separately.

The average indoor equivalent noise levels of the installation noises are shown in [Table 1].

The differences in equivalent A-weighted levels between blocks A and blocks B and C were small and in the range of 2 dB. The LCeq levels indoors were 19 to 20 dB higher than the LAeq levels.

The average values of third octave band levels and standard deviations for the measurements in block A and blocks B and C are shown in [Figure 2],[Figure 3]. The figures include recommended levels in Sweden for low frequency noise indoors (SOSFS 1996:7/E). The curve representing the normal hearing threshold is included for reference (ISO 389-7:1996).

With windows closed, the sound pressure levelsfrom installation noise did only slightly exceed the Swedish recommendations for low frequency noise. When the windows were slightly opened, the sound pressure levels exceeded the recommendations by about 10 dB SPL. The measurements suggest that the sound pressure levels were somewhat higher for blocks B and C as compared to block A. The standard deviations of the third octave band sound pressure levels between the flats in the same block were in the range of 2.6 to 4.8. Data indicated that the sound pressure levels in third octave bands were about 2-4 dB SPL higher on floor 5 compared to floor 1 to 3, but the measurements were too few to make firm conclusions about this finding.

The outdoor noise level from the installations was 57 dB LAeq 24h. The level during the day (07.00-19.00) was 58 dB LAeq, during the evening (19.00-23-00) 56 dB LAeq, and during the night (23.00-07.00) 57 dB LAeq.

Calculated noise levels outdoor and indoor from the traffic are shown in [Table 2].

The LAeq levels were about 10 dB higher for blocks B and C as compared to block A.

The maximum noise levels were similar for block A,B and C and amounted to about 78 dB outdoors and 50-51 dB indoors.

Noise annoyance

The major sources of annoyance indoors were reported for noise from fans/compressors, cooking smell from restaurants, exhaust from traffic and noise from traffic [Table 3]. As can be seen in [Table 3], more than 1/3 of the respondents reported that they were very or extremely annoyed indoors by noise from fans/compressors. Annoyance from cooking smells from restaurants was reported by about the same number of respondents as annoyance from traffic exhausts. Only three respondents (7.7%) were annoyed by music from neighbours and only two (5.4%) were annoyed by impact noise from neighbours.

The data were divided according to persons who had bedrooms facing the courtyard (CY) and persons with bedrooms facing a street (ST). [Figure 4] shows the proportion of subjects very or extremely annoyed by traffic exhaust, noise from traffic and fan/installations. Annoyance from traffic exhaust and traffic noise was mainly reported by the ST sample, while noise from fan/compressors was reported by 44% among the CY sample and 26% among the ST sample. It can also be noted that the prevalence of annoyance in the ST sample caused by noise from fans/compressors was equivalent to the reported annoyance from traffic noise, although their apartments had both living rooms and bedrooms facing the street. There was no significant difference in annoyance from fans/compressors between the ST and the CY samples (z=-0.965, p=0.36), while the annoyance caused by traffic noise was significantly higher in the ST sample (z=-4.630, p Sleep disturbance

The answers to the most important sleep related questions divided into the ST sample and the CY sample are shown in [Table 4]. The questions on sleep quality, difficulty in falling asleep and feelings of being tired and irritated in the morning ranged from 1 to 5, where 1 indicates a positive value ("very good sleep quality", "not at all difficult to fall asleep" and "very alert/rested" and "very relaxed" in the morning") and 5 represents a negative value ("very poor sleep quality", "difficult to fall asleep every day" and "very tired" and "very tensed" in the morning).

The table also shows that the ratings of the sleep related parameters were similar among respondents with bedrooms facing the street vs bedrooms facing the courtyard. A significant difference was found for the reports of feeling tired in the morning, however, where more subjects in the ST sample reported being tired (z=-2.316, pMethods

In epidemiological terms, the study population is very small and does not include a referent group. The results must thus be valued accordingly. The emphasis of the data presentation is therefore descriptive, and special caution must be used in interpreting the statistical analysis of e.g. sleep data, as no adjustment has been made for factors that can influence the sleep parameters.

This study is the first part of a series of studies, which will be followed up by further studies among similar populations in the city of Goteborg and will include referent groups not exposed to low frequency installation noise. In most cases, we will also be able to measure the response after noise abatement programs aimed to reduce the installation noise have been undertaken. At the end of this three-year programme, more valid data will thus be gained.

The subjective responses were obtained by questionnaires masked as a general living environment study. This was done in order not to direct the respondent's attention to noise, which could result in a more prominent annoyance response. Before the study was undertaken, two noise complaints had been reported to the health and environmental authorities, but there were no widespread complaint actions. The answers to the questions on general and specific items dealing with the general living environment and the conditions of the flats indicate that the attempt to mask the real purpose of the questionnaire was successful.

Noise measurements were made in a sample of flats at different locations in the building. A more extensive measurement program would probably have yielded better precision concerning the variation in noise exposure, especially between different floors. The standard deviations were however comparable to previous studies of low frequency ventilation noise involving a more extensive measurement program (Persson Waye and Rylander, 2001).

Results

The results showed that the proportion of respondents who were annoyed by noise from fans/compressors was high. Annoyance caused by installation noise was reported indoors as well as when time was spent in the courtyard. Compared to annoyance caused by traffic noise, the proportion of respondents reporting annoyance caused by installation noise was more than twice as high. Similar differences were reported regarding disturbed rest and relaxation. When the data were subdivided according to residents whose bedrooms faced the courtyard and those whose bedrooms faced the street, the same pattern was found, although it was more pronounced.

The comparatively high annoyance reported for installation noises could be explained by a difference in the perceived necessity of the noise from the different sources. While noise from traffic may be considered as unavoidable in the urban environment, noise from installations that service mainly shops and restaurants may be considered as easier to avoid and hence more unnecessary. The judged unnecessity of neighbourhood noise and street noise has previously been found to be related to annoyance in a survey of 200 subjects living in an urban area (Graeven, 1975). The more unnecessary the noise was perceived to be, the greater noise annoyance was reported. Interestingly, the reverse was found for noise at work, the more necessary the noise was perceived to be, the greater annoyance was reported. However in another occupational study, where subjects were asked to judge their opinion of the possibilities to reduce the noise level at their work place, it was found that persons who thought it was possible to reduce the noise i.e. meaning that they thought it was unnecessary loud, were more annoyed than others (Kjellberg et al., 1996). In further studies, subjects' perception of the necessity of the noise sources should be recorded.

Another explanation of the high occurrence of noise annoyance from installations could be that tenants with traffic on one side of the flat feel that it is very important to be able to open one window without being exposed to exhausts or noise. The finding in this study that respondents with bedrooms facing the street reported about the same extent of annoyance from installation noise as for traffic noise seems to support that theory. The results from this study do however not lend themselves to any conclusions of whether the access to a quiet side would be beneficial from a health point of view, and there is today very little data to support the hypothesis, originally suggested by Kihlman (1993). A recent paper indicates however, that the access to a quiet side resulted in overall lower noise annoyance and also less perceived noise induced sleep disturbances as compared to living in a flat where both sides face a trafficked road (Kihlman et al., 2002). However the authors regard the results as preliminary and hence more data is needed.

The character of the installation noise is another probable explanation for the annoyance response. The noise has a monotonous, continuous character and the low frequency character of the installation noise will penetrate into the flat and increase the annoyance potential. While intermittent noise is usually reported as more annoying and more disturbing of sleep (e.g. Ohrstrom and Rylander, 1982), less is known of the effects of long term exposure to continuous noise and of noises with a low frequency character. A recently performed experimental sleep study showed that a low frequency ventilation noise at 40 dBA, led to a significant increase of time to fall asleep and an altered cortisol response in the morning (Persson Waye et al., 2003). The results of that study and this field study thus point to the relevance to further study the annoyance potential of installation noises with a large proportion of low frequencies.

The study also points to the importance of obtaining a better basis for effects of low frequency noise at different levels and to consider low frequency noise outdoors as an important factor for indoor effects. The indoor levels with closed windows were at or slightly above the recommended levels for low frequencies indoors (SOSFS 1996:7/E). The high prevalence of annoyance indoors suggest that it is not only the noise levels indoors with closed windows that are of relevance for noise annoyance indoors, but also the noise levels that occur indoors when the windows are open.

The significant relation between greater annoyance and floor level was an unexpected finding. The increase of annoyance is supported by the indication that the noise levels were somewhat higher on the higher floor levels. This is in contrast to traffic noise measurements, where a reduction of the level is found with greater height of the building. The increase in level could be explained by the fact that some of the sources were located on the roof of the buildings but could also be a result of the shape of the enclosed area of the courtyard that probably acts as a resonator for the low frequencies.

Conclusion

The high disturbance of installation noises found in this study indicates the importance of also regulating the noise exposure on the "quiet side" of buildings in order to achieve a reasonable living environment. This is believed to be especially important for residents exposed to high noise levels of traffic noise on the other side of their dwellings. Further studies will give a better base for the extent of annoyance and acceptable levels of installation noises.[21]

References

1	Altman D.G. (1991) Practical statistics for medical research. Chapman and Hall, London.
2	Berglund B., Hassme´n P. and Job R. F. S. (1996). Sources and effects of low frequency noise. J. Acoust. Soc. Am. 99: 2985-3002.
3	Bryan M.E. (1976). Low frequency noise annoyance. In Infrasound and Low Frequency Vibration. Tempest W. ed., Academic Press Inc., London LTD, pp65-96.
4	Challis L.A. and Challis A.M. (1978). Low frequency noise problems from gas turbine power stations. In Proceedings of the International Conference on Noise Control Engineering. Lang W.W. ed., San Francisco, USA, pp475-480.
5	Chatterton P.F. (1979). A case history of a low frequency noise problem. Noise Control Vibration Isolation, AugustSeptember: 295-298.
6	Cocchi A., Fausti P. and Piva S. (1992). Experimental characterization of the low frequency noise annoyance arising from industrial plants. J. Low Freq. Noise Vibr. 11: 124-132.
7	Fields J. M., De Jong R.G., Gjestland T., Flindell I.H., Job R.F.S., Kurra S., Lercher P., Vallet M., Yano T., Guski R., Felscher-Suhr U. and Schumer R.(2001). Standardized general-purpose noise reaction questions for community noise surveys: research and a recommendation. J. Sound Vib. 242: 641-679.
8	Graeven D. (1975). Necessity, control, and predictablitiy of noise as determinants of noise annoyance. J. Soc. Psychol. 95: 85-90.
9	ISO 389-7:1996. Acoustics Reference zero for the calibration of audiometric equipment. Part 7: Reference threshold of hearing under free-field and diffuse-field listening conditions.
10	Jonasson H. and Nielsen H.L. (1996). Road Traffic Noise - Nordic Prediction Method, ISBN 92 9120 836 1, Tema Nord 1996:25, Nordic Council of Ministers.
11	Kihlman T. (1993). Action plan against noise. SOU 1993:65. ISBN 91-38-13397-0 (in Swedish).
12	Kihlman T., Ohrstrom E. and Skanberg A. Adverse health effects of noise and the value of access to quitness in residential areas. In: Proceedings of Inter Noise 2002, The 2002 International Congress and Exposition on Noise Control Engineering, Dearborn MI USA, August 19-21, 2002. Paper 484
13	Kjellberg A., Landstrom U., Tezars M., Soderberg L. and Akerlund E. (1996).The effects of nonphysical noise characteristics, ongoing task and noise sensitivity on annoyance and distraction due to noise at work. J. Environ. Psychol. 16: 123-136.
14	Kjellberg A., Muhr P. and Skoldstrom B. (1998). Fatigue after work in noise - an epidemiological survey study and three quasi-experimental field studies. Noise and Health 1: 47-55.
15	Mirowska M. (1998). An investigation and assessment of annoyance of low frequency noise in dwellings. J. Low Freq. Noise Vibr. 17: 119-126.
16	Persson Waye K. and Rylander R. (2001). The extent of annoyance and long term effects among persons exposed to low frequency noise in the home environment. J. Sound Vib. 240: 483-497.
17	Persson Waye K., Clow A., Edwards S., Hucklebridge F. and Rylander R. (2003). Effects of night-time low frequency noise on the cortisol response to awakening and subjective sleep quality. Life Sciences. 72: 863-875.
18	SOSFS 1996:7/E. General Guidelines issued by the Swedish National board of Health and Welfare. Adopted 15th May 1996. Indoor Noise and High Sound Levels.
19	SP-REPORT 1996:10. Measurement of low frequency sound in rooms. Swedish National Testing and Research Institute, Boras, Sweden.
20	Verzini A.M., Frassoni C.A. and Skarp A.H.O. (1999). A field research about effects of low frequency noises on man. Acta Acustica 85, suppl 1: S16.
21	Ohrstrom E. and Rylander R. (1982). Sleep disturbance effects of traffic noise - a laboratory study on after effects. J. Sound Vib. 84: 87-103.