Noise Health Home 

[Download PDF]
Year : 2004  |  Volume : 6  |  Issue : 24  |  Page : 1--7

Stress effects of noise in a field experiment in comparison to reactions to short term noise exposure in the laboratory.

H Ising1, R Michalak2,  
1 Umweltbundesamt a.D., Berlin, Germany
2 Fliman Hospital, Haifa, Israel

Correspondence Address:
H Ising
Rheinstr. 69, D-14612 Falkensee


Reactions to noise-induced communication disturbance of 42 men during a seminar were investigated. Stress reactions with or without road traffic noise (L m = 60 dB(A)) were compared. Traffic noise was played back via loudspeakers during one day in the seminar room. The following parameters were measured: Fatigue and mental tension by questionnaire; blood pressure and heart rate; excretion of adrenaline, noradrenaline and cAMP from the collected urine. The same subjects participated in a laboratory test where the blood pressure was measured during 5 minutes of rest and after 5 minutes of exposure to intermittent white noise (L m =97 dB(A)). It was found that the noise in the field experiment caused psychological and physiological stress effects in half of the subjects. Increased mental tension was correlated to increases as well as decreases of the blood pressure. Systolic blood pressure reactions were stronger than the reactions of diastolic blood pressure. Noise sensitive subjects reacted stronger than the others. In the short-term laboratory test, systolic blood pressure increases were smaller than the diastolic increases. At the end of the 5 minutes noise exposure only the diastolic blood pressure increases were significant. There was no correlation between the blood pressure reactions in the two different noise exposure experiments. There existed a positive correlation between noise sensitivity and the systolic blood pressure increases during the seminar, whilst the correlation, between noise sensitivity and systolic blood pressure increases in the laboratory exposure, was negative. From these results we conclude that short-term noise exposure experiments do not provide information about the effects of long-term real life exposure to environmental noise. Potential health effects of chronic noise-induced disturbances of activities are discussed.

How to cite this article:
Ising H, Michalak R. Stress effects of noise in a field experiment in comparison to reactions to short term noise exposure in the laboratory. Noise Health 2004;6:1-7

How to cite this URL:
Ising H, Michalak R. Stress effects of noise in a field experiment in comparison to reactions to short term noise exposure in the laboratory. Noise Health [serial online] 2004 [cited 2023 Feb 3 ];6:1-7
Available from:

Full Text

 Introduction and definition of the problem

The following quotations serve to define noise and noise effects (Jansen and Klosterkotter, 1980). "Noise is not a physical parameter; a sound event is perceived as noise if the impact it makes is detrimental to a person's somatic, mental and/or social comfort." "The type and intensity of noise effects may depend on:

the physical properties of the sound events;

the characteristics, attitudes and activities of the person affected;

the features of the situation in which sound affects the person."

Such noise effects can best be studied in field experiments in real life situations with experimentally altered noise exposure. Most of the laboratory studies, however, have only concentrated on sound with varying physical properties. This type of experiment deals with direct, vegetative effects of noise without cortical influence. Indirect noise effects, however, are caused by noise-induced disturbances to various activities, provoking different types of cortical responses, including psychological stress reactions such as nervous tension, annoyance etc. which are usually accompanied by somatic stress reactions.

Lehmann and Tamm (1956) already distinguished between these two types of reactions to noise. Indirect noise induced stress effects can occur at much lower noise levels than direct vegetative noise effects. Quite often they last for the whole duration of the noise induced disturbance whereas direct effects are mainly short term reactions to changes of noise exposure. For a review of the research on noise induced stress effects see (Ising & Braun, 2000).

The aim of the present paper is to compare somatic effects caused by noise induced disturbance of communication and effects of short term exposure to intensive white noise in the same healthy subjects - in short, to compare indirect and direct noise effects. This comparison may deepen our understanding of acute non­specific stress effects of low level noise in real life situations and their potentially long-term consequences as risk factors for diverse diseases.

 Experimental studies

Field study

The field study took place at the Police College in Spandau, Berlin. The experiment was described to the participants in an advanced training course, whereupon 42 persons volunteered to take part in the study. Some of those police officers, who described themselves as highly sensitive to noise, did not participate. It is likely, therefore, that selection was biased towards people who are less sensitive to noise.

The 42 test subjects were men aged 24 to 54 (average age: 41). All 42 underwent a thorough medical examination to establish their general state of health. The following medical parameters were measured and recorded: medical check-up, family and personal medical history, blood and urine analysis, blood pressure, ECG, pulmonary function test, thorax X-ray, drugs used. Although some medical parameters were found to be outside the normal range, none of the 42 candidates had to be excluded from the study on medical grounds. Additionally, the noise sensitivity of each subject was determined by questionnaire.

Method of field study

In order to exclude adaptation effects, the field study was carried out during the second week of the 14-day training course. The 42 subjects were divided at random into two groups before undergoing the procedure. Each group was monitored for one day with, and one day without noise exposure. The experimental conditions were switched around for the two groups to prevent serial effects. Differences of the effect variables during these two days were interpreted as noise effects.

In addition to the effects of noise induced stress we studied effects of emotional stress. For this purpose the blood pressure of the subjects was monitored for one day during which a highly exciting emotional theme (the use of weapons by the police) was taught and discussed. During the control days and the days with noise exposure, emotionally neutral themes were taught and discussed (basic police law etc.).

For the field experiment a satisfactory speech intelligibility was intended. For this purpose road traffic noise with a mean sound level of L M = 60 dB (A)was chosen. Due to the Lombard effect with this background noise, the speaker had to raise his voice to about LS, 1m = 72 dB(A).

This is rated as a loud voice (ISO/DIS 9921, 1992). Within a distance of 3 - 6m from the speaker, the signal to noise ratio will be about 0 dB corresponding to an articulation index of AI = 0.4. This hearing situation can be rated as satisfactory (Lazarus et al. 1985; Lazarus, 1990). The signal to noise ratio of 0 dB resp. AI = 0.4 corresponds to a syllable comprehensibility of 55% - 60% and to a sentence comprehensibility of 98% (ISO/TR 4870, 1975; Kryter ,1970; Lazarus et al. 1985).

The following effect variables were studied: Psychological condition. The psychological condition of the subjects was surveyed at midday by means of a questionnaire. A distinction was made between two factors which contribute to describing activation according to Thayer (1978). One relates to the antithesis "fatigue versus physical fitness", and the other to "balanced condition versus mental tension".

The questionnaire contained a list of eight adjectives relating to each of the two factors in mixed sequence. The subjects were asked to describe how they were feeling on a scale of 1 to 5, for which Rohrmann (1978) had tested the equidistance of neighbouring categories.

Urine samples. Urine was collected from each subject during class to determine the following parameters: adrenaline, c-AMP and noradrenaline (for methods see Ising 1980).

Blood pressure. Blood pressure was measured with a semi-automatic instrument, (Stereocard manufactured by Dr. Lange, Berlin,) which had a switching device for four blood pressure cuffs. Korotkoff's sounds were monitored by oscilloscope, to exclude wrong readings of the semi-automatic device (Dienel, 1987). The systolic and diastolic blood pressure of each subject was measured five times during teaching hours (8 a.m. to 2.30 p.m.) at intervals of a little over an hour. Each value was constituted by the mean value of at least three readings.

Heart rate. Parallel to this, the semi-automatic Stereocard and an electronic mean-value calculator were used to record the heart rate five times, during the day's classes.

Method of the short term laboratory study

The test procedure was explained in detail to all the subjects beforehand. The tests took place in a quiet room - (background noise L m m = 97 dB(A). To reduce the likelihood of startle reactions, the noise stimulus was increased gradually over three seconds to full level. White noise was produced by the Random Noise Generator Type 1402 from Bruel & Kjaer and applied via headphones, Beyer DT 48.

As in the field study, the semi-automatic blood pressure gauge was used to measure systolic and diastolic blood pressure at the beginning and end of both the rest and the noise phase. Each blood pressure value was constituted by taking the mean from at least three readings. A general rest value was obtained by averaging the values for the beginning and end of the rest phase. This was compared with the values at the beginning and the end of the noise phase and the average of both of them was taken.

The statistical significance of mean values of individual differences for effect parameters measured under control and noisy conditions was assessed using the Wilcoxon rank test, which is indifferent to distribution. Spearman's rank correlation coefficient served to analyze links between changes in the various parameters.

Three levels of significance are indicated:

* : pField study results

Examples of verbal comments.

The test persons were asked how they felt during the day with noise exposure. A few typical verbal comments will serve to illustrate the basic attitude of the subjects:

"When the noise was suddenly switched on at 8 o'clock, I have to say that until I was used to it, about 20 minutes, it was very distracting. By midday I was used to it, and then it wasn't all that hard to follow the lesson. But after the midday break I had had enough; it was more difficult to concentration again."

"Above all the monotony, the same noises over and over again..., that was very disruptive in the afternoon."

"Sometimes the teacher was hard to understand."

"My head was buzzing by the end."

"Around half-way through the lessons I realized that it had given me a bit of a headache."

Fatigue and mental tension

The differences in group mean values for the parameters fatigue and mental tension with and without noise exposure were as follows:

Arousal: -0.9 *** and Mental tension: +0.6 on a five point scale (42 subjects).

Urine analyses

The only parameter of the urine analyses to reveal significant differences was noradrenaline (NA related to creatinine). The noise-induced differences were as follows:

ANA = +10.0%** (41 subjects).

Blood pressure

The blood pressure was averaged over the 6.5 hours of each seminar day.

Statistically significant blood pressure reactions were observed. Systolic blood pressure increases were stronger than diastolic increases (1.7* / 1.1* mmHg (n=42)).

The effect of emotional stress upon blood pressure (+ 2.0* / + 1.4* mmHg (n = 40)) was

Heart rate

No significant differences were observed in the heart rate on the noisy day as compared with the control day.

Correlation analyses

Rank correlation coefficients of changes in psychological, circulatory and biochemical parameters (noise effect parameters) observed during classroom exposure to traffic noise as compared with no noise are given in [Table 1].

It should be pointed out that there was no significant correlation between changes in tension and in systolic blood pressure. However, there was a statistically significant correlation between changes of psychological tension and the magnitude of the changes of systolic blood pressure ΔAp s*). Beside this, we found a correlation between the changes of noradrenaline and cyclic AMP and between the changes of the systolic blood pressure and that of the heart rate.

There was also a significant correlation between noise sensitivity and the changes of systolic blood pressure (ΔAp s)

ΔAp s = R = +0.40* (n = 41).

Results of the short-term laboratory study Blood pressure

The noise-induced blood pressure changes during the short-term noise exposure test were as follows (n = 41):

First minute of noise:

ΔAp s = 2.8*** mm Hg

ΔApd = 3.2*** mm Hg

Last minute of noise:

ΔAp s = 1.1 mm Hg (n.s.)

ΔApd = 3.1*** mm Hg

Compared with the level at the beginning of the noise phase, the systolic blood pressure at the end was significantly lower by 1.7* mm Hg. No evidence was found of a fall in diastolic blood pressure during the noise phase. In the laboratory study the diastolic blood pressure increase was more pronounced than the systolic increase.

In [Table 2] the blood pressure differences caused by emotional stress, noise stress and short term noise exposure are compared. A different tendency between the blood pressure reactions in the field experiments and the short term noise exposure was indicated by the more pronounced diastolic blood pressure increases during the short term noise exposure. This was confirmed by a Wilcoxon test between the individual blood pressure differences in these experiments. Significant differences existed between the diastolic blood pressure differences of both field experiments as compared to the laboratory experiment. There were no significant differences between the blood pressure changes under emotional and noise stress.

Correlation analyses

Between noise sensitivity and changes in systolic blood pressure in the first minute a significant negative correlation was found:

R= - 0.32* (n = 39).

The correlation between the blood pressure values during the rest period in the field study and the laboratory study (syst: R= 0.74***, diast: R= 0.70***) was quite high, indicating the validity of the method of blood pressure measurement. There was no significant correlation between the blood pressure changes in the field experiment and the short term noise experiment.


Methodological consideration and individual blood pressure changes

The experiments described here are centered on the study of noise-induced changes in blood pressure. An accuracy of ± 5mm Hg was assumed to allow for uncertainties of a single measurement. Since each time 16 measurements were taken on the day with noise and the control day, the methodological inaccuracy of the mean blood pressure of one person comes to ± 1.25 mmHg. In the laboratory study the inaccuracy lay between 2 and 3 mmHg due to the fewer number of single measurements, so that mean blood pressure differences between experimental conditions and control conditions of more than 4 mmHg can be counted as real effects.

Using this supposition, we found after 6.5 hours noise-induced communication disturbance 5 and 3 persons with blood pressure decreases, and 13 and 11 persons with blood pressure increases respectively (systolic resp.diastolic). During 6.5 hours of emotional stress we found 7 and 1 person with decreased and 13 and 10 persons with increased blood pressure respectively (systolic resp. diastolic). In the short-term experiment we found 3 and 1 blood pressure decreases and 13 and 14 blood pressure increases respectively (systolic resp. diastolic).

Comparison of blood pressure reactions under laboratory and field conditions

The noise-induced blood pressure reactions in the field experiment and in the laboratory showed important differences. In the laboratory, it was mainly the diastolic blood pressure which increased, whereas in the field mainly the systolic blood pressure increased. There was no significant correlation between the individual blood pressure reactions in the two experiments. The correlation analyses also showed that the systolic blood pressure increases in the field experiment were positively and in the laboratory they were negatively correlated to the noise sensitivity.

The main conclusion from these results is: ­Direct noise effects do not correlate with indirect noise effects. It is wrong therefore to extrapolate the effects of low level noise exposure in real life situations from the effects of high level noise exposure in the laboratory.

Individual differences of noise effects

Noise exposure for several hours in real life situations may result in no psychological or physiological reactions, if the individual is able to compensate for the noise effects. In this case no disturbance of activities is experienced. The ability to compensate for noise may be learned in a short time, after the beginning of the exposure. This ability may end after several hours when the reserves of the individual start to be exhausted (see verbal comments).

If a person is not able to compensate for noise effects he will experience noise-induced disturbance of activities which are followed by psychological and physiological stress reactions. In our field study we observed an increase of mental tension under noise exposure in 50% of the subjects and an increase of noradrenaline which indicates an increase of the sympathetic activity, which leads to an increase of the total peripheral resistance. The reaction of the blood pressure regulatory system to the increase of the peripheral resistance determines the blood pressure reaction:

1.) The blood pressure may remain constant if a decrease of blood flow (the product of stroke volume and heart rate) exactly balances the increase of the peripheral resistance.

2.) If the counter regulation leads to a more pronounced decrease of the blood flow than the increase of the peripheral resistance, a decrease of the blood pressure will result. This results in a decreased blood supply to parts of an organism which may have negative long term health effects.

5 / 3 subjects showed more than - 4 mmHg (averaged over 6.5 hours) for the systolic / diastolic blood pressure decreases, respectively, which were correlated to increases of the psychological tension. This type of person may develop long term negative health effects but there is certainly no increased risk of hypertension. If the counter effect of the blood pressure regulation system is less than the increase of the peripheral resistance, the blood pressure will increase. 18 / 11 subjects showed systolic / diastolic blood pressure increases, respectively, of more than 4 mmHg (averaged over 6.5 hours). This type of person may under chronic repetition of noise stress be liable to an increased risk of hypertension and other heart and circulatory diseases.

The relevance of noise-induced disturbances of various activities as precondition for indirect noise effects on blood pressure was demonstrated in the CORDIS study. Long term occupational noise exposure caused chronic blood pressure increases only on those workers, who performed complex jobs (Melamed et al, 1999). In this case, the noise exposure caused an increased effort to perform the job and this resulted after 2-4 years in increases of systolic and - to a lesser degree - diastolic blood pressure.

Besides the above mentioned blood pressure increases we observed also some small blood pressure decreases, which we interpret, too, as noise effects. This is in accordance with the clearer results of a patient study (Ising and Havestadt, 1983). In a group of 50 female patients, 12 reacted to traffic noise exposure for 6 hours with L m = 65 dB(A) with a mean systolic and diastolic blood pressure decrease of 19 and 8 mm Hg respectively. 9 patients showed the opposite effect: an average systolic and diastolic blood pressure increase of 22 and 8 mm Hg respectively as comparing to a control day with hourly blood pressure measurements as well as a control period of three weeks with daily blood pressure measurements. These blood pressure changes were shown to be statistically significant effects of noise-induced disturbance of recreation.


Traffic noise is experienced as a negative environmental factor particularly when it disturbs human activities (e.g. communication, concentration, recreation etc.). This causes mental stress reactions which are usually accompanied by biochemical and circulatory stress effects.

Short-term laboratory tests provide no information about the effects of long-term exposure to environmental noise.[13]


1Dienel D. (1987) Psychische, ergonomische, physiologische und biochemische Wirkungen von Verkehrslarm. Dissertation, Berlin.
2Ising, H., Braun, C. 2000 Acute and chronic endocrine effects of noise: review of the research conducted at the Institute for Water, Soil and Air Hygiene. Noise & Health 7: 7-24.
3Ising H, Havestadt C.H., (1983) GesundheitlicheAuswirkungen von Verkehrslarmbelastungen. Bundesgesundheitsblatt 26, 76 - 79.
4ISO/TR 4870: (1975)Acoustics - The construction and calibration of speech intelligibility, Internat. Standard Org. Geneva.
5ISO/DIS 9921: (1992) Ergonomic assessment of speech communication (SIL method) Internat. Standard Org. Geneva.
6Jansen G., Klosterkotter W. (1980) Larm und Larmwirkungen. Bundesministerium des Inneren, Bonn.
7Kryter K.D. ,(1970) The effects of noise on man Academic Press. New York, London.
8Lazarus H., Lazarus - Mainka, G. Schubeins M. (1985) Sprachliche Kommunikation unter Larm Kiehl Vlg. Ludwigshafen.
9Lazarus H. (1990) New methods for describing and assessing direct speech communication under disturbing conditions Environmental International 16, 373 - 392.
10Lehmann G. and Tamm J. (1956) Die Beeinflussung vegetativer Funktionen des Menschen durch Gerausche.Wirtschafts- und Verkehrsministerium Nordrhein-Westfalen. Diisseldorf 1956 (Forschungsberichte no. 257, 1-37)
11Melamed S. Kristal-Boneh E. Froon P. (1999) Industrial noise exposure and risk factors for cardiovascular disease: Findings from the CORDIS study. Noise & Health; 4, 49­ 56.
12Rohrmann B. (1978) Empirische Studien zur Entwicklung von Antwortskalen fur die sozialwissenschaftliceh Forschung Z Sozialpsychol . 9, 222-245.
13Thayer R.E. (1978) Factor analytic and reliability studies on the activation-deactivation adjective check list. Psychol Rep 42, 747-756.