Noise Health Home 

[Download PDF]
Year : 2002  |  Volume : 4  |  Issue : 16  |  Page : 39--46

Cortical excitations, cortisol excretion and estimation of tolerable nightly over-flights

Manfred Spreng 
 Dept. Physiology and Experimental Pathophysiology, University of Erlangen, Germany

Correspondence Address:
Manfred Spreng
Dept. Physiology and Experimental Pathopysiology, University of Erlangen, Universitaetsstrasse 17, D-91054 Erlangen


Noise induces cortisol excretion even below the awakening threshold. This is based upon the existence of very close subcortical central nervous connections between parts of the auditory system (e. g. amygdala) showing typical plasticity effects and the hypothalmic-pituitary­adrenal (HPA)-axis. Repeated noise events (e.g. over-flights during nigh-time) will lead to accumulation of the cortisol concentration in blood. This happens because its time constant of exponential decrease is about 50 to 10 times larger than that one for adrenaline and noradrenaline. A twofold attempt has been made to calculate the cortisol accumulation using an initial value of noise induced small cortisol increase (rounded value 14 ng/ml) at the nightly threshold of beginning vegetative overreaction around 53 dB(A). A mean time-constant of 64 min has been applied based upon experimental studies. Using in a first step the range of minimal and maximal normal cortisol values as border line and taking into account a relation between peak sound pressure level and cortical excitation given by a power function (exponent 0.32, based on evoked potential studies in man) results in a formula to estimate tolerable events during night-time periods (over-flights in a given time range). Examples of results for 8 hours in the night are for instance values of 11 events with 55 dB(A) indoor peak level or 5 events with 75 dB(A) indoor peak level respectively. Those values of tolerable nightly noise events estimated on the basis of physiological processes and peak levels cannot be recalculated as or compared with equivalent sound levels.

How to cite this article:
Spreng M. Cortical excitations, cortisol excretion and estimation of tolerable nightly over-flights.Noise Health 2002;4:39-46

How to cite this URL:
Spreng M. Cortical excitations, cortisol excretion and estimation of tolerable nightly over-flights. Noise Health [serial online] 2002 [cited 2022 May 19 ];4:39-46
Available from:

Full Text


Noise exposure, even with quite low levels and below the wakening threshold, has the potential to cause stress hormone releases in humans. Frequent exposure may be linked with hormone increases leading to health problems in the long run (For recent reviews see Babisch, 2000; Ising and Braun, 2000; Spreng, 2000a). This is due to the close connection between the lateral amygdala as part of the auditory processing system -our subcortically permanently awake and most important warning system- and the hypothalamus via the central amygdala (Spreng, 2000b, Ising and Prasher, 2000).

Repeated noise events, especially noise with remarkable onset slopes starting from a quiet background as caused for instance by nightly over-flights of civilian aircraft, thus can lead to an accumulating increase of stress hormones.

This accumulation may reach quite high and even abnormal values if the stress hormone cortisol is regarded being metabolized with a time constant in the range of 60 minutes or more (Adrenaline: seconds to 3 minutes; Noradrenaline 7 to 12 minutes).

Therefore it might be of interest to think about a model based on cortisol accumulation which can help to estimate the number of tolerable noise events during night time, especially nightly over­flights.

Raw estimation of tolerable nightly events (over-flights) based upon accumulation of cortisol level

Although the stress hormone releases are regulated by complex feedback mechanisms and subdue to circadian rhythms [Figure 1] presents a strongly simplified scheme neglecting those complexities in a very first approximation.

[Figure 1] shows the accumulation of cortisol level starting with the first noise event at a mean value c m of plasma concentration. The noise event at the beginning of the first night hour increases the concentration by the initial value c i

In this very first step we neglect also the exponential increase of the cortisol release induced by the single noise events and take into account only the exponential decrease (drawn linear in the sketch of [Figure 1] to simplify matters).

Thus, assuming equidistant noise events over night time T (e.g. 22.00 to 06.00 or 00.00 to 04.00 hours) the accumulation of the initial increases c i produced by each event will reach higher and higher values and may even exceed a tolerable value c tol of plasma concentration.

In a selected night time interval T the number n of tolerable events to avoid the exceeding of the tolerable plasma concentration c tol is given by the formula [1] shown below, with τ as time constant of the cortisol decrease.


In the following plausible values for the parameters τ,, c m , c tol and c i are discussed and a possible relation between c i and the peak levels L max of the sound events is established.

Time constant τ of plasma cortisol decrease

To find out a reasonable value of the time constant τ of the decreasing time course of mean cortisol plasma concentration findings of experimental studies in human subjects were used (Gotthardt et al., 1995).

In 20 normal controls of two different age groups (mean 34 and 69 years of age) during and after to a cognitive challenge paradigm the time course of cortisol response had been measured in intervals of 15 minutes. In the young controls the mean cortisol values changed from around 40 ng/ml to 80 ng/ml and decreased with a time constant of approx. 60 minutes. In the older controls a change from around 40 ng/ml up to 120 ng/ml was observed decreasing with a time constant of approx. 69 minutes.

Therefore a mean value of

τ= 64 minutes

has been chosen for the time constant τ in the model under consideration.

Mean (c m ) and tolerable (c tol ) value of plasma cortisol concentration

Because of the circadian variability of the plasma cortisol level controlled by the HPA­system such a thing as a clear base-line or a quasi-constant mean value c m cannot be defined.

Therefore to simplify the model in the first step we can look for long time mean values of plasma cortisol levels published in the literature. In the same way it is possible to find out tolerable values (normal maximal values) in humans.

Thus typical values for c m and c tol are reported as listed in [Table 1].

Therefore the difference range (possibly riding on the circadian baseline)

c tol - c m needed in the model in a first step can be arranged as

c tol - c m = 231 ng/ml - 140 ng/ml = 91 ng/ml.

Basal value c i0 of the single evoked cortisol releases c i

To answer the question how to find a basal value c i0 of the single initial cortisol releases c i caused by the noise events during night time is not easy. At the moment there exist two possibilities to get a plausible estimation for such a value at low peak levels of noise (nearby a peak level which can be defined as threshold of vegetative overreaction as done below)

Firstly: Estimation of basal initial cortisol release c i0 from experimental data.

In their experiments using aircraft noise of lower peak level [e.g.55 dB(A)] in the sleeping rooms Maschke et al. (1995) found a significant accumulation of cortisol in the morning urine of the sleepers.

The authors used 16 aircraft noise events (over­flight noise) presented equidistant during a 4 hour period (every 15 minutes) in the night and measured around 1.5 hours after this series an increase of the total cortisol concentration in the urine corresponding to 109 µg/24h up to 131 µg/24h (difference 22 µg/24h). Based upon descriptions in the literature a mean relation between plasma cortisol concentration and cortisol concentration in the urine exists as 2.05 to 1 (Geigy, 1968: Urine 70 mg/24h corresponds to plasma 144 ng/ml).

Therefore a total difference of plasma concentration in the experiment mentioned above can be calculated to Δ = 45 ng/ml ( 22 * 2.05) after 4 hours stimulation plus an additional decrease during 1.5 hours.

Using the time constant of the exponential cortisol decrease mentioned above a value of the basal cortisol release can be estimated as

Δc = 45 ng/ml = (15 * c i0 * e -15/64 + c i0 ) * e -90/64


c i0 = 14.28 ng/ml

can be calculated as result of this estimation , still under the assumption that peak levels of 55 dB(A) are close to the threshold of vegetative overreaction during night time (see below).

Secondly: Estimation of basal initial cortisol release c i0 from normal episodic fluctuations.

Quite another approach to find out a basal value c i0 of the single evoked cortisol releases c i is given by analysing the spontaneous oscillations of the episodic changes of the plasma cortisol level under normal conditions. Those normal spontaneous oscillations superimposing the circadian rhythm of the baseline of the plasma cortisol level [Figure 2]a as reported for instance by Voigt (2000) show a large variability. Nevertheless they can be measured and averaged to a mean peak-to-peak amplitude of 28.84 +/­23.4 ng/ml.

Defining the positive half-wave as equal to the basal value of the single evoked cortisol releases results in

c i0 = 14.42 ng/ml.

This quantity corresponds rather good to that one found above (14.28 ng/ml evaluated out of the experimental data) and therefore a value rounded off to

c i0 = 14 ng/ml

will be used in the model under consideration.

Connection of the initial cortisol releases c i to the indoor peak levels L max of the sound events (over-flights)

Remembering the very close nervous connections between the auditory system and the hypothalamic-pituitary-adrenal axis (HPA-axis) via the amygdala (Spreng, 2000b), in a first approximation we are qualified to assume a direct spreading of the central nervous excitations E caused by noisy sounds.

Within the auditory system the activation produced by stimulating sounds of different intensity is represented in the size of evoked potentials objectively measurable in human subjects.

This activation (the amplitudes of the evoked responses) is closely related to the central nervous excitation E, which can be regarded as a kind of nervous energy level.

As shown with [Figure 2]b there exists an intensity function between sound pressure level and relative amplitudes of evoked potentials following a power function (linear relation in the double-logarithmic scale of [Figure 2]b) with an mean exponent k = 0.32 (Spreng, 1979). That exponent is close to well-known findings of Stevens (1961) using instantaneous (closely connected to the central nervous excitations) handgrip scaling of sensation of sound intensity. His experiments not only result in relations being described as power functions, but also in an exponent of 0.35 for auditory stimulation.

This leads to a relation between sound intensity and central nervous excitation as

E/E0 = (I/I 0 ) k

and therefore to

c i /c i0 = (I/I0) k k = 0.32 (2)

Introducing the sound intensity level L = 10 lg I/I 0 equation (2) changes to


and finally to


where L is the peak sound pressure level L max of the indoor noise event and L 0 a reference level as defined below.

The number n of tolerable noise events (over­flights) may then be generally calculated by combining equations (1) and (4) according to


or L max in relation to the number n (n > 2) of nightly events according to


Definition of a reference level L 0 as a threshold of beginning vegetative overreaction during night-time

There exists a good agreement (e.g. Rovekamp, 1983, Jansen et al,. 1999) that significantly measurable vegetative reactions (threshold of vegetative overreactions ?) caused by sounds during day-time starts between L max = 60 and 65 dB(A).

By the way these values corresponds well to levels of a loud human voice at a distance of 2 to 4 meters (L max = 61 to 69 dB) and thus may have developmental background.

In very early electrophysiological studies (Keidel and Spreng, 1976) we found in human subjects

- deviations from the normal intensity function of the auditory evoked potentials short after preceding noise stimulation - the quick EEG-Alpharhythm (13 Hz) reaching the same amplitudes as the slow EEG­Alpharhythm (9 Hz)

- significant changes of the psychogalvanic skin response when the peak values of the sounds (or noise bursts) exceed the mean value of L max = 63 dB(A).

On the other side it is well known (DiNisi et al., 1990; Jansen et al. 1995) that during night time the organism is more sensitive to sounds (about 10 to 15 dB lower threshold).

Thus we can define some kind of nightly threshold of vegetative overreaction at least at a peak level of

L 0 = 53 dB(A).

This definition is supported by the findings of measurable stress hormone releases caused by aircraft noise with peak levels of 55 dB(A) during sleep (Maschke et al., 1995) and reductions of duration of deep sleep, total sleep, REM-sleep etc. caused by multiple (e.g. 50 events) peak levels in the range 42 to 55 dB(A)[Griefahn, 1990, Ohrstrom, 1995, Vallet and Vernet, 1993].

Calculation of tolerable nightly over-flights with different peak levels

Using all the values elaborated in this paper [ τ = 64 minutes, c tol - cm = 91 ng/ml, c i0 = 14 ng/ml, k = 0.32 and L 0 = 53 dB(A)] for T = 8 hours night time equation (5) is calculated numerically according to


or L max in relation to the number n (n > 2) of nightly events (equation [6]) according to


The resulting tolerable number n of nightly events (over-flights) is plotted in [Figure 3] as bar chart with plausible limiting values and the exact relation is presented with [Figure 4].

For example it is readable from this [Figure 3] that 11 events with L max = 55 dB(A) indoor peak level or 5 events with L max = 75 dB(A) indoor peak level are just tolerable.

It must be emphasised that those values of tolerable nightly noise events estimated on the basis of physiological processes and peak levels cannot be recalculated as or compared with equivalent sound levels.


As far as we know up to now there exists no clear relation between tolerable maximal indoor levels (L max ) and the number of noise events during night time (e. g. nightly over-flights of civilian aircraft).

Based on experimental sleep studies in the laboratory Griefahn (1990) developed a relation connecting maximal peak levels, number of noise events, and the probability to wake up. This has led to border lines between awakening reactions and no reactions. The resulting curves fall off rapidly for frequencies of 1 to 5 flights per night and show a similar shape as the curve presented here [Figure 3],[Figure 4].

However, because of the completely different approach detailed comparisons, especially statements in the range of high peak levels, are not possible.

Indeed, based on a 10% risk of awakenings Griefahn (1990) proposed a critical peak noise level of L max = 53 to 54 dB(A) for 10 to 30 flights per night, the lower number being in agreement with the data presented here [11 to 13 over-flights with peak levels around L max = 50 to 55 dB(A)].

Close to a conclusion of Vallet and Vernet (1993) the extrapolation of our model [Figure 3] beyond 20 to 25 occurrences per night results that aircrafts then need to be very quiet [indoor levels below L max = 40 dB(A) for sleepers with open windows].

Although the simplificative model and the data presented here are based upon some preconditions which can be discussed in more detail, the resulting relation between number of tolerable nightly noise events and maximal indoor peak levels L max at least has two remarkable and plausible corner-stones.

Firstly: The highest indoor peak level L max = 115 to 120 dB(A) for one to two single over­flights (mean duration 20 seconds) per night is close to the limit of SEL = 125 dB(A) to absolutely prevent hearing impairment as found reviewing animal experiments which show permanent morphological damage (Spreng, 1990).

Secondly: The correspondence to findings of Maschke et al. (1995) resulting increased cortisol levels (45 ng/ml near to c tol - c m /2) against controls after 16 over-flights during only half the night with indoor peak levels of L max = 55 dB(A) instead of the tolerable 8 events for that level (T = 4*60 min = 240 min in equation 5); and the rough agreement with the statement of Griefahn (1990) allowing L max = 53 dB(A) for 10 flights per night.[23]


1Babisch, W. (2000) Traffic noise and cardiovascular disease: Epidemiological review and synthesis. Noise & Health, 8: 9-32
2Binoux, M., Girard, F., Mozziconacci, P. (1965) Variation in the plasma corticotropin and cortisol activities during the day and night and under influence of metopirone and ACTH. Ann. Pediatr. Paris 12(6): 466-474
3DiNisi, J., Muzet, A., Ehrhart, J., Libert, J. P. (1990) Comparison of cardiovascular responses to noise during waking and sleeping in humans. Sleep 13: 108-120
4Golenhofen, K. (2000) Physiologie heute. Urban & Schwarzenberg, Munchen-Wien-Baltimore, pp. 413-114
5Gotthardt, U., Schweiger, U., Fahrenberg, J., Lauer, C. J., Holsboer, F., Heuser, I. (1995) Cortisol, ACTH, and cardiovascular response to a cognitive challenge paradigm in aging and depression. Am. J. Physiol. 268:R865-R873
6Griefahn, B. (1990) Praventivmedizinische Vorschlage fur den nachtlichen Schallschutz. Zeitschrift fur Larmbekampfung 37: 7-14
7Hollmann, W., Hettinger, T. (2000) Sportmedizin, Schattauer Verlag, Stuttgart- New York, pp. 87-90
8Ising, 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
9Ising, H., Prasher, D. (2000) Noise as a stressor and its impact on health. Noise & Health, 7: 5-6
10Jansen, G., Linnemeier, A., Nitzsche, M. (1995) Methodenkritische Uberlegungen und Empfehlungen zur Bewertung von Nachtfluglarm. Zeitschrift fur Larmbekampfung, 42: 91-106
11Jansen G., Notbohm, G. Schwarze, S. (1999) Gesundheitsbeeintrachtigungen durch umweltbedingten Larm. In Umwelt und Gesundheit, Rat der Sachverstandigen fur Umweltfragen, eds. Metzler­Poeschel, Stuttgart, pp. 249-343
12Keidel W. D., Spreng, M. (1976) Neuro-elektrophysiolo­gische Larmbewertung. Report BMI/UB II 5-520-01, Technische Informations Bibliothek, AC 2890, Hannover.
13Liddle, G. W. (1966) Analysis of circadian rhythms in human adrenocortical secretory activity. Arch. Intern. Med. 117(&): 739-743
14Maschke, C., Arndt, D., Ising, H., Laude, G., Thierfelder, W., Contzen,S. (1995)Nachtfluglarmwirkungen auf Anwohner. Schriftenreihe des Vereins fur Wasser-, Boden­und Lufthygiene, Gustav Fischer Verlag, Stuttgart-Jena­New York
15Ohrstrom, E. (1995) Effects of low level s of road traffic noise during the night: A laboratory study on number of events, maximum noise levels and noise sensitivity. Journal of Sound and Vibration 179, 603-615
16Rovekamp, A. J. M. (1983) Physiological effects of environmental noise on normal and more sound-sensitive human beings. In Noise as a Public Health Problem, Vol 1, Rossi, G,ed. Edizione Tecniche (Amplifon), Milano, pp. 605-614
17Spreng, M. (1979) Improvement of ERA by speech­specific stimulation and correction of amplitude and latency behaviour. In Evoked Potentials. Barber, C. ed. TP Press Ltd, Lancester pp. 329-336
18Spreng, M. (1990) Effects of noise from military low-level flights on humans: Part I In Noise as a Public Health Problem, Vol 4. Berglund, B.; Lindvall, T. eds. Swedish Council for Building Research, Stockholm, pp. 293-303
19Spreng, M. (2000a ) Possible health effects of noise induced cortisol increase. Noise & Health 7:59-63
20Spreng, M. (2000b) Central nervous system activation by noise. Noise & Health 7: 49-57
21Stevens, S.S. (1961) The psychophysics of sensory function. In Sensory communication. Rosenblith,W. A. ed. MIT Press and John Wiley & Sons, Inc. New York­London, pp. 1-33
22Vallet, M., Vernet, I. (1993) Night aircraft noise index and sleep research results. In Noise and Disease, Ising,H., Kruppa, B. eds. Gustav Fischer Verlag, Stuttgart-New York, pp 413-415
23Voigt, K. (1996) Endokrines System, In Lehrbuch der Physiologie, Klinke, R., Silbernagel, S. eds. Georg Thieme Verlag, Stuttgart-New York, pp. 467-468