Artigos

Introduction
Occupational noise is the most common and preventable cause of disabling hearing loss in adults. Noise-induced hearing loss (NIHL) is the second most common occupational disease1 and is related to mechanical, metabolic and cellular changes.2-3
The etiopathogenesis is multifactorial, with a genetic component, exposure to noise, chemicals, among others.1
The role of diabetes mellitus (DM) in hearing loss has been investigated since the 50's, being the subject of debates and controversies until today.4 According to the Diabetes Federation (2019 - 2020), there are about 425 million adults with DM in the world, and the projection indicates that in 2045, there will be a 48% increase in this number in the adult population. The prevalence of the disease is related to a fast urbanization, with the consequent adoption of a sedentary lifestyle and a diet rich in carbohydrates, and to the aging of the population.5
Among the complications of DM, peripheral polyneuropathy stands out, which is the result of changes in nerve cells triggered by glycemic decompensation. Hyperglycemia, if sustained over a long period, can induce cranial nerve damage, progressing from malnutrition to membrane necrosis and nerve cell demyelination. This degeneration can lead to a reduction in the conductivity of the electrical impulse, and consequently, in hearing sensitivity.6-7-10-11-12
Once installed, most cases of hearing loss can be irreversible, especially if it has a sensorineural origin.13
NIHL appears initially and predominantly at higher frequencies (3kHz, 4kHz and 6kHz). Although the frequency of audiometric tests established by Brazilian legislation be semiannual, few has been studied about the contributing factors for the development of hearing loss among workers occupationally exposed to noise.8 The high prevalence of DM in the economically active population points to the relevance of investigating its influence on the development and worsening of NIHL.
The objective of this study was to evaluate the association between noise exposure and diabetes mellitus in the early onset of occupational hearing loss in individuals exposed to occupational noise from different work environments and different intensities

Methods
Retrospective cohort study, conducted among active workers, of both genders, aged between 19 and 59 years old, exposed to occupational noise, employed in sugar-ethanol plant in the region known as Nova Alta Paulista, countryside of São Paulo, from 2010 to 2020. All workers with a record of admission audiometric examination and recent periodic examination were included.
This study was approved by the Instituto de Assistência ao Servidor Público Estadual de São Paulo – IAMSPE Research Ethics Committee, with assent procedure number 3,631,114. All workers exposed to noise above 70 dB (A) who underwent audiometry at the company's Occupational Health Clinic were invited to participate in the study by the main researcher. They were informed that occupational records data would be collected from their medical records, without the possibility of identifying the subjects, according to the research project described in the Free and Informed Consent Form. After acceptance, these workers signed the Form and received a copy.







Data collection
The data used followed a 12-month interval between the reference and sequential. For data collection, a structured and pre-coded form was used, containing personal background and medical history, medication use, occupational data, and work history with emphasis on exposure to noisy environments, non-occupational noise exposure, use of safety equipment, impression of one's own hearing, and significant audiological backgrounds.
The audiometric examination was performed to determine the hearing thresholds of workers exposed to high levels of sound pressure and to elucidate the diagnosis of hearing loss. However, this is an exam that directly depends on the patient's response. Several precautions were taken regarding the examination to guarantee its quality and trustworthiness. The audiometric examination was preceded by a prior meatoscopy carried out by the professional responsible for conducting the examination, to check for earwax plugs or any foreign body, and individuals were asked for a minimum acoustic rest of 14 hours prior to the examination, so that effects such as Temporary Threshold Shift (TTS) did not occur in order to obtain a more reliable result.
Air conduction frequencies of 250, 500, 1000, 2000, 3000, 4000, 6000, and 8000 Hz were tested and, when altered, bone conduction testing was included at 500, 1000, 2000, 3000, and 4000 Hz. Audiometric examinations were conducted in an acoustic environment where sound pressure levels did not exceed international recommendations (ANSI 3.1 (1991) or OSHA 81 Appendix D parameter). Audiograms were classified according to the audiometric tracing defined by Ordinance 19 (04/09/98), Annex I of Regulatory Norm 7 of the Ministry of Labor, and described as normal tracing for individuals with hearing thresholds up to 25 dBHL, suggestive tracing of Noise Induced Hearing Loss (SPAINPSE) for individuals with auditory alterations at frequencies of 3000 and/or 4000 and/or 6000 Hz, of the sensorineural type; or in cases where auditory alteration was verified at other frequencies, the thresholds of 3000, 4000, and 6000 Hz will necessarily be more depressed, and non-suggestive tracing of Noise Induced Hearing Loss (NSPAINPSE) for examinations with auditory alterations whose results do not fit into any of the previous definitions.
In the natural history of Diabetes Mellitus, pathophysiological changes are present before glycemic values are above reference values, but still below the diagnostic values for Diabetes Mellitus, called pre-diabetes. Insulin resistance is already present and, in the absence of measures to combat modifiable risk factors, it often evolves into clinically manifested disease. The diagnosis of diabetes can be made through laboratory tests as: fasting blood glucose, oral glucose tolerance test (OGTT) and glycated hemoglobin (HbA1c). Confirming the diagnosis of the disease requires repeating the altered tests. The results were considered in percentage (%) and the individual was considered normoglycemic, pre-diabetic or diabetic.
Occupational records were reviewed to obtain demographic, clinical, and noise exposure information.
The dependent or outcome variable was the presence of sensorineural hearing loss, defined by the decrease recorded in the last exam in relation to the admission exam. The independent or predictive variables were: age (grouped according to the expected physiological change in hearing), environmental sound pressure level (categorized into risk classes for hearing loss), type of workplace (agricultural or industrial) and the presence of DM (referred or diagnosed during periodic examinations).



Statistical analysis
The estimated sample size was 365 patients, considering 2 standard deviations for confidence level, 70% of the phenomenon percentage, 5% maximum error, and infinite population. The statistical analysis was performed in stages:
1. Comparison of the relative frequencies of NIHL for each of the studied variables categories. In the comparison, the probability function adopted for the hypothesis test was Pearson's chi-square.
2. In this stage, factors with greater significance were identified. Subsequently, univariate Cox proportional hazards regression was applied. The Cox proportional hazards model provides predictors of death with their probability of occurrence (risk) compared to the reference group, which presents the lowest risk for death. The adequacy of the model is verified by the likelihood ratio test (log likelihood).
3. Variables that were identified in the univariate analysis as significant (p<0.05) or that have an influence on NIHL (age, occupational sound pressure level and work environment) were included in the multivariate logistic regression model. For this, the “stepwise forward” method was used, in which each variable was inserted until the best fit for the maximum likelihood value was obtained. The statistical program Stata 15.1 (StataCorp LLC) was used for the analyses.


Results
In the period from 2010 to 2020, 547 individuals were evaluated. The age ranged from 19 to 59 years old (mean 35±8 years), with the majority of workers engaged in an agricultural environment (64.2%). The prevalence of diabetes among the workers in the sample was 5.9% (CI 4.15% - 8.17%)
Analyzing Table 1, it is observed that statistically significant differences were found for the presence of DM (p<0.001) and for the age group to which the worker belongs (p<0.001).
The univariate evaluation of the NIHL predictors obtained from the logistic regression model, showed that the chance of finding individuals aged over 39 years old among workers with hearing loss was 4.25 times greater than the chance of finding workers aged between 19 and 28 years old with the same condition. Regarding the presence of DM, the chance of having individuals with DM among workers with hearing loss was 4.63 times greater than the chance of finding individuals without the disease (Table 2).
In the multivariate model, when adjusting for the other predictor variables, there is a reduction in the Odss Ratio (OR) for the presence of DM, since part of the observed effect would be associated with the worker's age (Table 3).

Discussion
The World Health Organization (WHO) estimates that 10% of the world population would be exposed to sound pressure levels that can potentially lead to NIHL, which makes occupational noise exposure a priority public health problem.8
Hearing losses are caused by multiple factors and may be related to age, occupation, years of noise exposure, lack of use of individual hearing protection equipment, exposure to chemicals, infectious and degenerative diseases.
NIHL can even be confused with presbycusis, and studies are highly divergent, with references that can vary its onset between 45 and 65 years of age. The differential diagnosis between these two types of loss can be made with certainty by analyzing the audiogram, which shows a lesser hearing decline in NIHL.
The present study indicates an association between NIHL and DM among workers at a sugar and ethanol plant. Individuals exposed in their workplaces are susceptible to the development of hearing loss, with its early onset and accelerated evolution due to the presence of this comorbidity that leads to an hearing structure degeneration. This loss can be avoided by actions for the control of noise exposure in the work environment, as well as health care for the prevention, early diagnosis and control of DM.
It would be advisable to review the legislation regulating labor laws regarding noise exposure and the mandatory use of ear protection for only individuals exposed to noises above 85 dB, and to include examinations for investigating diabetes diagnosis in pre-employment medical examinations.
The increased risk for individuals over 39 years old can be partly explained by the correlation between the development of DM and the individual's aging, since it is a chronic-degenerative disease. Furthermore, the difference found in the comparison with younger individuals may reflect an occupational past in environments with low noise control.
Another relevant finding in this study was the decreased risk for workers exposed to sound pressure levels above 85 dBA when compared to those exposed to lower levels. This result can be explained by the mandatory use of personal protective equipment by workers who are exposed to noise above 85 dBA.12 The association found could be investigated through etiological studies, to analyze the safety of the threshold currently adopted in the prevention of NIHL.
The cross-sectional nature of this study does not make possible to study the occupational noise-DM interaction in the etiopathogenesis of hearing loss, since the reverse association could also be an interpretation for the model obtained. Workers exposed to sound pressure levels above 85dBA would be continuously subjected to a stressor that would induce the maintenance of higher glycemic levels. Consequently, the risk for the development of DM or decompensation of pre-existing disease would be higher for this population.

Conclusion
In conclusion, the findings of this study suggest an association between DM and the onset of hearing loss in individuals exposed to occupational noise below 85 dB in a sugarcane mill. In individuals exposed to noise levels above 85 dB, auditory protection appears to be effective in preventing hearing loss.





















References
1.Lie A, Skogstad M, Johannessen HA, Tynes T, Mehlum IS, Nordby KC, et al.Occupational noise exposure ande hearing: a systematic review.Int Arch Occupm Environ Health., 2016;89(3):351-372.
2.Nemati S, Hassanzadeh R, Mehrdad M, Sajedi KS. Hearing Status in Patients with Type 2 Diabetes Mellitus According to Blood-Sugar Control: A Comparative Study. Iran J Otorhinolaryngol., 2018;30(99):209-218.
3.Bernardi APA, (org). Conhecimentos essenciais para atuar bem em empresas: Audiologia Ocupacional. São José dos Campos: Editora Parma, 2003.
4.Amaral MAS, Assencio FVA. Perda Auditiva em Pacientes Diabéticos. Revista Cefac., 2002;4(7):71-76.
5.Diretrizes da Sociedade Brasileira de Diabetes 2019-2020. São Paulo: Editora Clannad, 2019.
6.Jorge LG, Sandra PS, Joíza LC, Angela JR, Mirela JDA. Diabetes Melito: Diagnóstico, Classificação e Avaliação do Controle Glicêmico. Arq. Bras. Endocrinol. Metab., 2001;46(1):16-26.
7.Meneses-Barriviera CL, Bazoni JA, Doi MY, Marchiori LLM. Probable Association of Hearing Loss, Hypertension and Diabetes Mellitus in the Elderly. Int Arch Otorhinolaryngol., 2018;22(4):337-341. doi:10.1055/s-0037-1606644
8.Oishi N, Schacht J. Emerging treatments for noise-induced hearing loss. Expert Opin Emerg Drugs., 2011;16(2):235-245.
9.Cavalcanti EL. Efeitos auditivos e extra - auditivos relacionados à exposição ao ruído em trabalhadores com perda auditiva induzda por ruído ocupacional em uma usina sucroalcooleira. Bahia: UFBA, 2020. http://repositorio.ufba.br/ri/handle/ri/31711.
10.Przewo?ny T, Gójska-Grymaj?o A, Kwarciany M, G?secki D, Narkiewicz K. Hypertension and cochlear hearing loss. Blood Press. 2015;24(4):199-205., doi:10.3109/08037051.2015.1049466.
11.Wattamwar K, Qian ZJ, Otter J, et al. Association of Cardiovascular Comorbidities With Hearing Loss in the Older Old. JAMA Otolaryngol Head Neck Surg. 2018;144(7):623-629., doi:10.1001/jamaoto.2018.0643.
12.Tikka C, Verbeek JH, Kateman E, Morata TC, Dreschler WA, Ferrite S. Interventions to prevent occupational noise-induced hearing loss. Cochrane Database Syst Rev., 2017;7(7):CD006396. doi:10.1002/14651858.CD006396.pub4
13.Wang TC, Chang TY, Tyler R, Lin YJ, Liang WM, Shau YW, et al. Noise Induced Hearing Loss and Tinnitus—New Research Developments and Remaining Gaps in Disease Assessment, Treatment, and Prevention. Brain Sciences. 2020;10(10):732., https://doi.org/10.3390/brainsci10100732