Dose - response relationship between noise exposure and the risk of occupational injury.

Many workers worldwide experience fatality and disability caused by occupational injuries. This study examined the relationship between noise exposure and occupational injuries at factories in Korea. A total of 1790 factories located in northern Gyeonggi Province, Korea was evaluated. The time-weighted average levels of dust and noise exposure were taken from Workplace Exposure Assessment data. Apart occupational injuries, sports events, traffic accidents, and other accidents occurring outside workplaces were excluded. The incidences of occupational injury in each factory were calculated by data from the Korea Workers' Compensation and Welfare Services. Workplaces were classified according to the incidence of any occupational injuries (incident or nonincident workplaces, respectively). Workplace dust exposure was classified as <1 or ≥ 1 mg/m³ , and noise exposure as <80, 80-89, or >90 dB. Workplaces with high noise exposure were significantly associated with being incident workplaces, whereas workplaces with high dust exposure were not. The odds ratios (95% confidence intervals) derived from a logistic regression model were 1.68 (1.27-2.24) and 3.42 (2.26-5.17) at 80-89 dB and ≥ 90 dB versus <80 dB. These associations remained significant when in a separate analysis according to high or low dust exposure level. Noise exposure increases the risk of occupational injury in the workplace. Furthermore, the risk of occupational injury increases with noise exposure level in a dose-response relationship. Therefore, strategies for reducing noise exposure level are required to decrease the risk of occupational injury.

Introduction

Many workers worldwide experience fatality and disability caused by occupational injuries.[1] Accordingly, accident causation models for understanding and preventing occupational injury have been proposed. The domino theory of Heinrich[2] is the most commonly used model; Heinrich suggests that injuries occur as a result of multiple errors involving an unsafe environment and unsafe behaviors causing the actual accident.

The components of unsafe environments include hazardous factors such as moving machines during work.[3] The absence of safety guards is a very simple hazard often faced by workers operating machines or equipment.[45] If unsafe behaviors occur in any unsafe environment, the risk of occupational injury increases drastically. The components of human unsafe factors include a lack of motivation, knowledge, training, performance, and safety awareness about accidents.[6]

Some studies suggest that environmental stress such as heat[7] as well as job stress[8] and fatigue[9] are associated with the risk of occupational injury by making workers less aware of safety. The hearing as well as visual impairment can increase the risk of occupational injuries or accident by increasing the human error due to lack of recognition of warning the signal.[10] Human errors also can be occurred in various workplaces without permanent impairment. For example, dust is generated in various workplaces and could be linked to human stress via eye irritation.[11] Excessive noise exposure can disrupt detection of warning signals or trigger fatigue and stress, which are related to safety unawareness and human error as mentioned above.[12] In the workplace of Europe, because the occupational noise exposure was too loud, almost one-third of workers should raise voices to keep a normal conversation[1314] while those conversations are important to recognize the warning signals.

In Asia, almost half of all workers in manufacturing are exposed to noise, and 30-50% of factories have high noise levels exceeding 90 dB.[15] The nonauditory effects of noise include increased sleeplessness and human errors, which are also possible mechanisms linking noise and occupational injury.[16] In the same context of that biological plausibility, there were some investigation about noise exposure and risk of occupational injury,[171819] but there is lack of evidences combination exposure of dust and noise on the risk of occupational injury. Therefore, elucidating the association between occupational injury and noise or dust exposure are important topics in occupational safety. The present study assessed 1790 factories to elucidate the associations of noise and dust with occupational injury.

Methods

Ethics statement

This study used a number of incidents where an injury has occurred in each factory without information about individual cases of workers. The injury cases information of individual workers were de-identified prior to count the injury incident number of each factory. The factory name and other records were also anonymized and de-identified prior to analysis. This study design was approved by the Institutional Review Boards of Yonsei University.

Factories

A total of 1990 factories located in northern Gyeonggi Province, Korea was included in this study. Factories that had no information regarding Workplace Exposure Assessment (WEA) data were excluded. Finally, the data of 1790 factories were analyzed.

Exposure assessments

The WEA is performed more than once per year by certified agencies from the Korea Ministry of Employment and Labor. All WEAs are performed by a certified industrial hygienist. WEAs target various hazardous agents including chemical compounds such as organic solvents, metals, acids, alkalis, gases, and metal working fluids as well as and physical agents like noise, heat, radiation, etc. Noise and dust exposures were measured by personal sampling methods using by dosimeters. One worker per every different job sectors in factories was chosen for personal sampling. After that, the highest value among personal samplings was used to calculate the 8-h time-weighted average (TWA). Because eye irritation occurs as a result of dust exposure exceeding 1 mg/m3,[20] a factory was defined as being a high-exposure factory when the TWA of dust exceeded this level. Meanwhile, noise levels of <80, 80-89, and ≥90 dB were categorized as negligible, marginal, and high exposure, respectively.[21]

Occupational injuries

Data from the Korea Workers’ Compensation and Welfare Service were used to determine the incidence of occupational injury at each factory in 2010. Injuries that occurred outside the factory, such as business trips by car, injuries from athletic events, and accidents due to natural disasters, were excluded. Among a total of 1790 factories, 375 factories recorded one or more occupational injuries; these factories were defined as “incident workplaces.” Meanwhile, factories not recording any occupational injuries were defined as “nonincident workplaces.”

Statistical analysis

The χ2 test was used to compare differences between incident and nonincident workplaces. The odds ratios (ORs) and 95% confidence intervals (95% CIs) of workplace incidences were calculated using a multiple logistic regression model. Model I was adjusted for factory size, Model II was adjusted for factory size and work pattern (i.e., daytime or shift work), and Model III was adjusted for factory size, work pattern, and dust exposure. Two-tailed P < 0.05 were considered statistically significant.

Results

Basic characteristics of factories

Nonincident workplaces included a significantly greater proportion of small factories (i.e., <30 workers) than incident workplaces in Table 1 (78.59% vs. 60.27%, P < 0.001). There was no difference between nonincident and incident workplaces with respect to the proportion of shift work. Incident workplaces had a significantly greater proportion of high noise exposure (i.e., ≥90 dB) than nonincident workplaces (14.67% vs. 6.93%, P < 0.001). There was no significant difference between nonincident and incident workplaces with respect to dust exposure level.

Table 1

Basic characteristics of factories according to risk of occupational injury

Occupational characteristics	Nonincident workplace	Incident workplace	P
Number of workers in the factory
<30 workers	1112 (78.59)	226 (60.27)	<0.001
30-49 workers	172 (12.16)	68 (18.13)
≥50 workers	131 (9.26)	81 (21.60)
Work pattern
Daytime work	1266 (89.47)	325 (86.67)	0.125
Shift work	149 (10.53)	50 (13.33)
Noise exposure (TWA)
<80 dB	492 (34.77)	76 (20.27)	<0.001
80-89 dB	825 (58.30)	244 (65.07)
≥90 dB	98 (6.93)	55 (14.67)
Dust exposure (TWA)
<1 mg/m3	1084 (76.61)	297 (79.20)	0.288
≥1 mg/m3	331 (23.39)	78 (20.80)

TWA = 8-h time-weighted average exposure level

Association between noise exposure and occupational injury according to dust exposure level

In a separate analysis according to a dust exposure cutoff of <1 mg/m3, the incident workplaces had a significantly greater proportion high noise exposure (i.e., ≥90 dB) than nonincident workplaces (16.50% vs. 7.47%, P < 0.001) [Table 2]. This association remained significant with a dust exposure cutoff of ≥1 mg/m3 (7.69% vs. 5.14%, P = 0.034).

Table 2

Association between noise exposure and incident workplaces according to dust exposure level

Dust exposure	Noise exposure	Nonincident workplace	Incident workplace	P
<1 mg/m3	<80 dB	399 (36.81)	65 (21.89)	<0.001
	80-89 dB	604 (55.72)	183 (61.62)
	≥90 dB	81 (7.47)	49 (16.50)
≥1 mg/m3	<80 dB	93 (28.10)	11 (14.10)	0.034
	80-89 dB	221 (66.77)	61 (78.21)
	≥90 dB	17 (5.14)	6 (7.69)

Odds ratios of incident workplaces according to high noise exposure

In crude logistic regression analysis [Figure 1], the ORs (95% CIs) of a factory with a noise exposure level of 80-90 and ≥90 dB being an incident workplace were 1.92 (1.45-2.54) and 3.63 (2.41-5.47), respectively, compared to a factory with a noise exposure level <80 dB [Figure 1, Table 3]. In factories with dust exposure level <1 mg/m3, the OR (95% CI) of a factory being incident workplaces with a noise exposure level of 80-90 and ≥90 dB was 1.86 (1.36-2.54) and 3.71 (2.39-5.77), respectively, compared to a factory with a noise exposure level of <80 dB; in factories with a dust exposure level ≥1 mg/m3, the respective ORs (95% CIs) were 2.33 (1.18-4.64) and 2.98 (0.97-9.16).

Figure 1

Odds ratios and 95% confidence intervals for occupational injury

Table 3

Multivariate logistic regression models of the association between occupational injury and noise exposure

Hazard	Model I	Model II	Model III
Noise exposure
<80 dB	1	1	1
80-89 dB	1.68 (1.27-2.24)	1.72 (1.29-2.30)	1.75 (1.31-2.33)
≥90 dB	3.42 (2.26-5.17)	3.49 (2.30-5.29)	3.42 (2.26-5.20)
Noise exposure (dust exposure <1 mg/m3)
<80 dB	1	1
80-89 dB	1.65 (1.21-2.27)	1.72 (1.25-2.37)
≥90 dB	3.57 (2.28-5.57)	3.68 (2.35-5.78)
Noise exposure (dust exposure ≥1 mg/m3)
<80 dB	1	1
80-89 dB	2.04 (1.01-4.11)	2.03 (1.01-4.10)
≥90 dB	2.53 (0.81-7.89)	2.50 (0.80-7.80)

Model I: Adjusted for factory size, Model II: Model I + adjusted for work pattern (i.e., shift work), Model III: Model II + adjusted for dust exposure

The significant association between high noise exposure and incident workplace status remained significant even after adjusting for factory size and work pattern in Model II (OR [95% CI]: 1.72 [1.29-2.30] at 80-89 dB and 3.49 [2.30-5.29] at ≥90 dB). The association also remained significant after adjusting for dust exposure in Model III (OR [95% CI]: 1.75 [1.31-2.33] at 80-89 dB and 3.42 [2.26-5.20] at ≥90 dB).

In a separate analysis according to dust exposure level (i.e., <1 and ≥1 mg/m3, respectively), the association between noise exposure level and incident workplace remained significant: 1.72 (1.25-2.37) and 2.03 (1.01-4.10) at 80-89 dB and 3.68 (2.35-5.78) and 2.50 (0.80-7.80) at ≥90 dB.

Discussion

The results of the present study indicate that high noise exposure in workplaces is significantly associated with the workplace being an incident workplace (i.e., a workplace where occupational injuries occur). Furthermore, the results indicate a dose — response relationship between noise exposure and the risk of occupational injury. Our results support the other previous results reported from the western country[171819] to report an association between noise exposure and risk of occupational injury. This study is the first investigation to extend the association between noise exposure and the risk of occupational injury in Asia country. Noise exposure related to occupational injury independently to dust exposure, and there were also no combined effects of dust and noise exposure in the current study.

Noise is the most frequently reported occupational hazard in various workplaces.[14] Noise exposure is linked with various diseases such as hearing loss, cardiovascular disease, and psychological stress.[222324] Furthermore, excessive noise exposure can increase the frequency of unsafe behaviors by triggering the fatigue, stress,[16] and reducing amount of attention.[23] Hence, noise decreases safety and increases unsafe behaviors, consequently increasing the risk of occupational injury. Workers with noise induce hearing loss require higher dB warning signal to recognized the sound compare to healthy workers. Because noise exposure worker should wear the hearing protection equipment and warning signals can be drowned out by background noise in unsafe workplaces, noise can exacerbate the unsafe environment by making workers less sensitive to recognize of warnings. Actually, a well-designed retrospective cohort study observed 240,000 person-years reported that over the 12% of occupational accidents are due to a combination effect of noise-induced hearing loss (NIHL) and workplace noise exposure.[19] Some report also confirms the association between noise exposure and risk of occupational injury using by hospital admission data.[17] Our current results, which were derived from workers’ compensation data, also supported and extended that association by elucidating dose — response manner. Hence, the present results suggest noise exposure increases the risk of occupational injury possibly by increasing unsafe behaviors within already unsafe environments.

Noise-induced hearing loss could lead to stress and fatigue because workers needed the additional efforts to consent to understand speech-in-noise. Furthermore, NIHL workers had increased anxiety associated with the fear about might missing important information for their safety. Studies involving workers with hearing defects demonstrate that workplaces with high noise exposure (e.g., above 100 dBA) were associated with high risks of occupational injury.[1217] However, in Korea, workers who had noise-induced hearing loss with more than average 30 dB loss should be transferred to the workplace with noise <80 dB. In the current study, almost all workers working at the workplace with noise >80 dB noise have no NIHL. Hence, we cannot investigate the interaction between noise exposure and status of NIHL. In addition, other investigation reported that even relatively low level of noise exposure (e.g., 80 dBA) also can disrupt the verbal communication which related to detection of warning-signal in both hearing loss and normal hearing workers, temporally.[25] Hence, even relatively low level of noise exposure also can trigger the stress and fatigues of normal hearing workers, and our current study show factories exposed to 80-89 dB also had increased risk of occupational injury compare the factories exposed to below 80 dB did.

Dust from welding fumes, iron oxide, and crystalline silica, as well as organic dust, are generated in various workplaces.[26] Excessive exposure to dust irritates the eyes and respiratory system, increasing unsafe behaviors.[11] Furthermore, excessive dust exposure can coincide with unsafe environments that include hazardous machinery. However, there was no significant relationship between dust exposure and the risk of occupational injury in the present study. In this study, the risk of occupational injury was linked to noise exposure even after adjusting for dust exposure in the multivariate logistic regression analyses as well as a separate analysis, which may indicate they coincide in unsafe work environments. Hence, the association between noise exposure and occupational injury after adjusting for dust exposure suggests that noise exposure affects occupational injury after adjusting for unsafe environments.

The association between factory size and the risk of occupational injury is controversial: Inverse,[27] positive,[28] and U-shaped[29] associations have been reported. In the present study, there was a positive association between factory size and the risk of occupational injury. Therefore, the risk of occupational injuries in large factories may have been overestimated; consequently, the present results cannot be compared to those of previous studies that include the incidence rate of occupational injuries. Furthermore, almost 90% of factories included in the present study had <50 workers. Therefore, because the present results were based on small factories, the positive association between factory size and the risk of occupational injury cannot be generalized to larger factories.

This study has several limitations. First, because we used workplace average level of noise exposure and incident cases based on each factory, there was no exposure or injury information for individual workers. We have no information about personal protective equipment for noise and dust exposure. Hence, the workers in a given factory may have varying exposure levels; it is difficult to generalize the results to individual workers. Regardless, it is possible that the workers in the present study could share unsafe environments with their coworkers, because the factories were relatively small (i.e., <50 workers). Although different dust types have different toxicity and severity of irritation, we have no information of dust characteristic in the current study. We have no information about personal protective equipment for noise and dust exposure. Hence, more comprehensive exposure assessments were needed to clarify the association between dust exposure and risk of occupational injury. Furthermore, the dose-response effects of noise exposure in the multiple logistic regression analyses strengthen the relationship between noise exposure and the risk of occupational injury in the present study. In addition, safety education programs, such as training intervention of behavioral modification or provide guideline and information to enable the worker protect them self from unsafe environment, are factors associated with prevention strategy against to human unsafe behaviors,[30] while the systematic regulation of hazardous materials is associated with unsafe environments.[30] Furthermore, organizational and cultural factors are important components of unsafe environments that directly or indirectly influence unsafe behaviors.[31] Although working patterns (i.e., daytime and shift work) were controlled for in the analyses, other risk factors, such as workers age, length of experience in the specific work setting, industrial sector as well as organizational and cultural factors, were not included due to a lack of information. Because we used workers’ compensation data to detect incident of occupational injury, nondeclaration of injury incident was a possible problem in the current study. Therefore, further studies including comprehensive exposure assessment and injury detection strategy with examination of additional risk factors for occupational injury are required to elucidate the association between noise exposure and the risk of occupational injury.

Conclusion

The present study indicates that noise exposure increases the risk of occupational injury regardless of dust exposure level. Furthermore, there is a dose-response relationship between noise exposure and the risk of occupational injury. However, investigations including individual-level data are required to elucidate these findings.