INTRODUCTION

Noise does not affect only our hearing, although its effects are better perceived and characterized by this sense. Noise can also cause tinnitus, headaches, fullness in the ear, dizziness, gastric disorders (gastritis and gastroduodenal ulcer), transitory alterations in blood pressure, stress, vision impairment, lack of attention, memory loss, sleep disorders and mood changes.

The exposure-related disorders will depend on the frequence, intensity, noise duration and rhythm, as well as exposure time and individual susceptibility. Continuing noises are less harmful than broken sounds, because despite the fact that the first sound impact is received without protection, further sounds are damped by the ear protection mechanism. However, in broken noises the impact has no damping, because in between sounds the protection mechanism has time to relax.

The occupational hearing loss, or noise induced hearing loss (NIHL) is a hearing disorder that affects many workers exposed to noisy working environments and it may be fully preventable. Its diagnosis alone already points towards the failure of a whole preventive system that should have been made available to the worker.

Although it may be fully preventable, the rules for industrial noise control are not being followed. Thus, the worker may present important alterations that will impact his/her life quality, such as hearing impairment and the difficulty in understanding speech in noisy environments (TV, radio, movie theater, theater), or even sound signs such as emergency sirens, music and environmental sounds (CURADO et al, 2001).

Ferraz (1998) stated in his paper that the alterations in speech recognition that may occur after NIHL should not be seen only as hearing symptoms, because communication is of primary importance in social, professional and affective activities. Thus, verbal communication impairments may have a severe impact on the life quality of the individuals.

SELIGMAN (1993) described this effect on the life quality of individuals in relation to psycho-social and non-auditory factors, and the ones most commonly related to noise were: restlessness, anxiety, stress, fatigue, irritability, mood changes, isolation, solitude, sorrow, depression and negative self-image. These factors may appear alone or in association. The presence of a number of these symptoms leads us to a diagnosis of stress, which bears different impact on the individuals and their life quality.

In physics, stress means a force exerted on a material and tension means the very reaction of this material against the stress. Notwithstanding, when we try to transfer the knowledge of physics into pathology, it is not very clear whether the stress is related to the forces exerted on the individual by the external world, or reactions from this individual against these forces (response). Actually, when the term stress is applied to a living being, it is related to a mechanism that results from the individual-environment interaction and that aims at providing this individual with adaptive resources (PORTNOI & GLINA, 2002).

These resources are seen in face of what we call stressing factors, the stress causing agents, which may be biological, psychological (internal stress factors), social and/or environmental (external stress factors) in origin. Therefore, in order for something to become a stress factor, it is necessary that facts and/or situations be assessed as such and they should represent a change in the lives of the individuals in such a way as to involve adaptive efforts (PORTNOI & GLINA, 2002).

The reactions, or responses to stress (adaptation) are alterations and/or symptoms of organic, psychological and social origin that may be very uncomfortable. Factors such as intensity, frequence and duration of the stress factors, as well as the capacity and propensity of the individual to face them are essential in outcome characteristics (PORTNOI & GLINA, 2002).

The reactions accruing straight from the physiological process of stress may have consequences in the appearance and maintenance of diseases, such as reduction or enhancement of the immune response, hypo or hyper endocrine activity, alterations in the autonomic control balance, changes in sleep patterns and changes in brain neurotransmission, neuromodulation and neuroendocrine functions.

Lab experiments have proven that noise excess can alter the very size of many endocrine glands, it alters the respiratory rhythm and causes blood pressure variations. There have also been changes in heart rhythm and pupilar dilation.

Among secondary and side effects the following have been seen: increase in irritability, anxiety, nausea, fatigue and weariness, appetite loss, reduction in sexual activity and the appearance of a pre-neurotic state (SANTOS et al., 1992). Anyway, despite vast research and different stress-related manifestations, there is no consensus about what stress really is, nor its relationship with noise.
Some papers show some data regarding different populations exposed to occupational noise. MANIGLIA AND CARMO (1998) assessed the hearing of 15 joiners who worked at the Medical School of Rio Preto, exposed to noise that varied from 80 to 110 dB. They saw a 20% incidence of class I hearing alterations (Merluzzi), and verified the need for prevention programs in order to preserve the health of the workers.

MIRANDA et al. (1998) assessed the prevalence of noise induced hearing loss in factory workers of nine different industries in the metropolitan region of Salvador (BA). The work was carried out based on audiometric data from 7925 workers from 44 companies. All the workers assessed underwent at least one audiometric exam according to specific work legislation in effect: NR-7 - Ordinance 3214/78. Hearing losses were classified according to differentiation diagnostic criteria into sensorineural, conductive and mixed type hearing loss, broadly established by audiology and Noise Induced Hearing Loss (NIHL) patterns. Hearing thresholds were considered normal up to 25dB. Besides the type of loss, the audiometric curves were classified according to the degree of loss (MERLUZZI et al, 1989). Hearing loss prevalence was of 45.9% in the population considered. As to NIHL, adding both unilateral and bilateral hearing losses, there was a prevalence of 35.7%. The authors were surprised with NIHL in 18% of the workers evaluated. This study led to the recognition of one extremely alarming picture, pointing towards the great need for the companies to implement Hearing Conservation Programs.
ALMEIDA et al. (2000) assessed both the clinical and audiometric characteristics of noise induced occupational sensorineural hearing loss, according to age range and the exposure duration in years. 222 occupational sensorineural hearing loss patients were studied, and they correlated hearing complaint, audiometric threshold alterations in the 250Hz to the 8,000Hz frequency range, and speech recognition indexes versus the age range and the duration of exposure. As a control group, the audiometric threshold of a population of the same age, without a past history of hearing disease was used. The group was divided into subgroups, and three decades of exposure were analyzed. They saw that the clinical complaint of hearing loss increased with the increase in age and exposure duration, while the frequence of tinnitus complaint remained constant. The audiometric thresholds in the second decade of exposure presented variations that depended on the age range analyzed. The different audiometric curves achieved were parallel among themselves, not horizontal, and the worst thresholds were found in high frequencies: from 3,000Hz to 8,000Hz. Speech recognition also got worse as age and exposure duration increased.

Having in mind all the stress-related changes and their relationship with the exposure to intense noise, this paper aimed at analyzing the influence of daily exposure to intense noise levels in hearing loss in the social lives of joiners and locksmiths.

MATERIALS AND METHODS

This study was approved by the Research Ethics Committee and was developed at the Occupational Health Department of the University Hospital in the City of São Paulo, with people working as locksmiths or joiners for the hospital.
Table 1 depicts the demographic data of the population studied: age, function and exposure time. Of the 21 workers studied, 10 were locksmiths and 11 were joiners, with age varying between 31 and 67 years, with average age of 44.6 years. The minimum time they had been working in that position had been of 1 year and the maximum was 50 years, with an average of 19.95 years.

The sample selection criteria were: 1) Exposure to noise levels above 85 dBHL; 2) Auditory Rest for 14 hours before undergoing audiometry; 3) Reading and Signing the Informed Consent Form (Attachment I); 4) :Working in the same occupation for at least 1 year; 5) Maximum age of 60 years; 6) No conductive hearing loss, no past history of unilateral or bilateral ear diseases; 7) No unilateral or bilateral deep hearing loss.
The instruments used were pure-tone audiometry and the stress level assessment questionnaire (Attachment II).

All the participants were invited to participate in the present study through an invitational letter and had to sign the informed consent form (Attachment I) if they agreed to participate. After that, all participants were assessed by pure-tone audiometry, evaluating the frequency range between 250 and 8,000 kHz using a SD 25 (Siemens) audiometer in a soundproof booth. We collected data such as type of occupation, time spent in that occupation and age. We also had the participants answer a questionnaire about the effects stress had on their social lives, made up of 59 multiple choice questions (COVOLAN, 1989). Scores could vary between 0 and 177 points, thus representing the variation on stress level: the lower the score, the less stress, and vice-versa. To better understand the most affected aspects by noise, the questions were divided in somatic and psychosocial.

Audiometry results were classified according to COSTA (1988) and compared to the total stress score obtained. The Costa classification divides the audiometric curves according to each side in groups from 0 to 5 (Chart 2).

Group 0 has the audiograms in which hearing losses did not go beyond 25dB in all frequencies or in their averages. Group 5 had the curves corresponding to non-occupational hearing losses or technique artifacts. The noise induced hearing losses that presented characteristic drops in 3,000 and 6,000 Hz are then classified from 0 to 4. We then take two arithmetic averages of the losses in decibels. The former in 500, 1,000 and 2,000 Hz translates the quality of auditory discrimination within a soundproof booth. The latter translates the average in 3,000, 4,000 and 6,000 Hz and properly characterized the noise induced hearing loss. The loss in 3,000 Hz and later in 2,000 Hz translates the loss progression in terms of day-to-day intelligibility (COSTA, 1988).

Statistical Analysis
A statistical study of the results was carried out through the variables correlation analysis using the MINITAB software.
The data found were analyzed according to Pearson's correlation coefficient using p<0.05 as suggested in biological studies.
The correlation method used was that of Spearman's correlation coefficient r (non-parametric) and the Pearson's correlation coefficient. In order to do this, the data were ranked and the coefficient results lead us to infer on two hypothesis:

 Null Hypothesis: There is no correlation among the variables
 Alternative Hypothesis: There is correlation among the variables
Decision rules for the above hypothesis are:
 If p > 0.05, we do not reject the null hypothesis, in other words, there is no correlation among the variables.
 If p < 0.05, we reject the null hypothesis, in other words, there really is correlation among the variables.
 If the Pearson' correlation coefficient comes close to +1 or -1, there really is correlation among the variables, however, when this coefficient comes closer to zero, there is no such correlation. Usually, if the correlation coefficient is above 0.30, we consider that there is great correlation potential (it does not mean that such correlation exists), and if this value is above 0.50, then we may consider the correlation to exist.

RESULTS

The data obtained from each worker's audiometry are depicted on the second and third columns of Table 2. In cases of unilateral hearing loss, the individual was considered as carrier of hearing impairment; if there is different degrees of hearing loss for the same individual, the worst level was selected for this classification. The fourth column of Table 2 shows this classification.

We did not find levels 3 and 4 amongst our subjects and there was only one individual in level 5, the latter being a characteristic of losses other than NIHL.

Graph 1 shows the hearing loss characteristics of the population studied, where we can see that: within the hearing impaired population (48% of the sample), the majority presents level 1 hearing loss, according to COSTA (1988).

Graph 2 shows the number of individuals and the respective degree of hearing loss found in relation to their age range, where we notice a prevalence of hearing impaired individuals with ages varying between 41 and 50 years.

Graph 3 shows the degree of hearing loss related to the person's occupation. Level 1 hearing loss was found more frequently among joiners and level 2 among locksmiths.

Graph 4 shows the time, in years, that each employee has spent in the present occupation and the level of hearing loss found. Notice that the number of normal individuals is greater than the number of workers with some degree of hearing loss, regardless of time spent in the present occupation.

As to the answers given in the stress questionnaire - major goal of the present paper - the total score was compared to the time each worker spent in the present occupation (Graph 5) and with the level of hearing loss (Graph 6). Regardless of the time spent in that occupation, the greater number of answers varied between 0 and 20 points and for the different hearing levels (levels 0,1,2 and 5), most of the answers also varied from 0 to 20 points.

According to Table 3, the total questionnaire score was divided in two areas, namely: psychosocial reactions and somatic reactions. The greater number of answers was given to the psychosocial reactions. Notwithstanding, there were no significant correlation between each area and the hearing loss found (p = 0.866 and p = 0.202 respectively).

The questionnaire could have gotten to a maximum score of 177 points per person, and the highest score was 64 points (Table 3). Of the total score of points obtained by all the workers in all the questions, 72.4% was given to "never"; 21.2% to "a few times"; 3.7% to "often" and 2.7% to "always".

Statistical Analysis of the Results
The correlations between occupation, age, time in occupation, hearing loss level, total questionnaire score and score were divided in areas (Tables 4 and 5).

As shown in the above table, we may observe that there was no correlation among the variables. However, we may say that the correlation is potentially significant between the total score and the time spent in that occupation (0.630 > 0.30); and between age and total score (0.517 > 0.30).
There were no correlations between these variables. Each answer alone was correlated to the following variables: age, time in occupation, occupation and hearing loss. Following we mention only the statistically significant correlations.
The questions may be seen in Attachment II.
Hearing loss and question # 46 correlation ('I have dry mouth'):

Pearson's correlation: 0.438
P-Value: 0.047
There was a correlation between the variables: hearing loss and having dry mouth (P-Value < 0.05).

Correlation between age and question # 9 ('I have nightmares that cause anxiety'):
Pearson's correlation: 0.495
P-Value: 0.023

There was correlation between the variables: age and thoughts that cause anxiety (P-Value < 0.05).
Correlation between age and question # 34 ('I feel depressed'):
Pearson's correlation: 0.439
P-Value: 0.046

There was correlation between the variables: age and depression (P-Value < 0.05).
Correlation between time in occupation and question # 37 ("I feel tearing eyes and blurred vision"):
Pearson's correlation: 0.475
P-Value: 0.030
There was correlation between the variables: time in occupation and feeling tearing eyes and blurred vision (P-Value < 0.05).


DISCUSSION

The questionnaire used was initially proposed to assess the stress impact on a group of psychologists, that naturally had higher educational level than the population studied in this present research. Considering that this questionnaire may have words and expressions which are not known by the population investigated, some answers obtained may be questioned. In order to make the individuals at ease to answer the questionnaire without being intimidated by the presence of the investigator, the questionnaires were handed to the workers to answer individually and the investigator was made available to answer any doubts they could have. Few individuals required further explaining of words or expressions used in the questionnaire.

Analyzing the information asked by some workers, we could see that these and other questions were present among other workers, who for unknown reasons were reluctant in asking for further explanation to better answer the questionnaire. This would represent doubts in relation to the reliability of the results, for not being able to analyze if most of the answers given had been "never" for being true, for lack of knowledge about what was being investigated or for any other reason. Having said that, we may infer that the instrument used to assess the stress level of joiners and locksmiths was not adjusted for its complexity.

There was no significant correlation between the stress questionnaire score and the hearing loss, besides other correlations that were also not significant.

Some questions alone presented significant statistic correlation. Question # 46 presented statistically significant correlation with hearing loss. Questions # 9 and 34 significantly correlated to age, and question # 37 with time spent in the occupation. The correlation between hearing loss and dry mouth symptoms may be related to some stressful situation found in the person's life, however this was an isolated characteristic related to hearing loss and may not be considered as relevant among so many other issues approached that were not related to hearing loss.

The correlation between age and anxiety-causing thoughts, may be related with stress which may be a consequence of the individual's concern with the future and his/her family, since older persons have anxiety-causing thoughts more often than younger individuals.

The correlation between age and depression may be related to the future expectations on the person's life, and this may lead to a certain level of stress.

The correlation between the time in occupation and feeling tearing eyes with blurred vision may arise from the exposure to dust and air pollutants, having in mind how frequently this happens in these occupations. For having relationship with the time spent in the occupation, we may infer that the longer the exposure to these pollutants and dust the greater the risk of having some eye problem.

In spite of some isolated questions, the total questionnaire score presented potencialy significant correlation with age and time spent in the occupation, a statistically significant correlation possibility. With that we may infer that stress may be related to age and the time spent working in the same occupation.

In order to better see the answers obtained through the questionnaire, it was divided in psychosocial reaction and somatic reactions, so that we could check if any of these areas would be highlighted. This analysis did not show statistically significant correlation.

Through a clinical analysis, without statistical data, we could see that the most answers " a few times" (most answer given and one that indicated some symptom present) was acquired in the question about back pain (question 23), which if analyzed alone, may be related only to the very characteristics of these occupations, for they often times work in the upright position and do not have a comfortable position to perform his/her tasks. Therefore, one may question if this issue has any relation with stress itself, or if this is merely a consequence of bad posture and improper ergonomic characteristics of his/her tasks.
The "often" and "always" answers happened very few times. With "often", no question was answered by more than two people and with "always" the question that obtained the most answers (4) was related to the behavior of avoiding parties, games and social meetings, that alone also does not characterize any connection to stress.

SANTOS et al. (1992) and SELIGMAN (1993) stated that noise causes non-auditory effects such as: increase in irritability, restlessness, anxiety, tension, nauseas, insomnia, mood changes, stress, isolation, solitude, sorrow, depression, negative self-image, fatigue and tiredness, and also loss of appetite, sexual activity reduction and pre-neurotic state symptoms. However, the present study did not show significant correlation between noise and these symptoms. It is possible that these workers may not be exposed to loud enough sounds as to trigger auditory and psychosocial factors.

This may also be inferred due to findings related to the hearing loss found in this population which was of 48% in this sample, considering unilateral loss as a hearing loss. Only 4 individuals had hearing losses with thresholds above 35 dB, and that, often times, not even hearing factors may be perceived.
As previously mentioned, even the unilateral hearing loss individuals were considered hearing impaired, in order to facilitate the observation and the relationship between hearing loss and a possible manifestation of stress. However, only to illustrate, it is worth mentioning that this study presented asymmetrical losses in 14% of the subjects, similar to the MIRANDA et al (1998) findings, who found 18% of unilateral NIHL.
Common hearing loss-related symptoms, such as tinnitus (ALMEIDA et al., 2000), was not significantly mentioned in this study.

CONCLUSION

After analyzing and discussing the answers presented, we may conclude that, among joiners and locksmiths who work at the University of São Paulo Hospital:
 48% (forty eight) of the assessed population presents some degree of hearing loss;
 The hearing loss is still not perceived and does not cause social discomfort, thus not impacting the individual's life quality;
 There was no correlation between the degree of hearing loss, according to Costa (1988) and the stress level, according to COVOLAN (1989);
 There was a potentially significant correlation between age and the questionnaire score and between the time in occupation and the questionnaire score.

REFERENCES

Almeida SIC, Albernaz PLM, Zaia PA, Xavier OG, Karazawa EHI. História natural da perda auditiva ocupacional provocada por ruído. Rev Ass Méd Brasil, 46 (2): 143-58, 2000.
Costa EA. Classificação e quantificação das perdas auditivas em audiometrias industriais. Rev Bras Saúde Ocup, ; 61(16):35-38, 1988.
Covolan MA. O stress ocupacional do psicólogo clínico: seus sintomas, suas fontes e as estratégias utilizadas para controlá-lo. 1989. 115f. Dissertação (Mestrado) - Instituto de Psicologia da PUC de Campinas, Campinas.
Curado JAF, Rabelo WCSV, Alves W, Perini RF, Siqueira PH. A incidência da PAIR na lavanderia de um hospital universitário. Arq Otorrinolaringol, 5(2):33-36, 2001.
Ferraz NM. A questão da informação na conservação auditiva: a perspectiva do trabalhador portador de PAIR. Rev. Mundo Saúde, 22(5):291-7, 1998.
Maniglia JV, Carmo KC. Avaliação dos danos por ruído em trabalhadores de marcenaria. Acta Awho, 17(2):90-96, 1998.
Miranda CR, Dias CR, Pena PGL, Nobre LCC, Aquino R. Perda auditiva induzida pelo ruído em trabalhadores industriais da região metropolitana de Salvador, Bahia. Inf. Epidemiol. SUS, 7(1):87-94, 1998.
Portnoi AG, Glina DMR. Estresse no trabalho: abordagem individual e organizacional. Belo Horizonte, 2002, 42f (Monografia - Fundacentro. Centro Regional de Minas Gerais).
Santos MFC, Bethancourt MLBT, Tedesco MLF. Prevenção de problemas auditivos numa industria. Hosp Adm Saúde, 16(3):137-139, 1992.
Seligman J. Efeitos não auditivos e aspectos psicossociais no indivíduo submetido a ruído intenso. Rev Bras Otorrinolaringol, 59(9): 257-259, 1993.