Noise Health Home 

[Download PDF]
Year : 2002  |  Volume : 4  |  Issue : 16  |  Page : 23--31

The joint effect of industrial noise exposure and job complexity on all-cause mortality - The CORDIS study

Samuel Melamed1, Paul Froom2,  
1 Department of Occupational Health Psychology, National Institute of Occupational & Environmental Health,Raanana, Israel
2 Department of Epidemiology, National Institute of Occupational & Environmental Health, Raanana, Israel

Correspondence Address:
Samuel Melamed
Department of Occupational Health Psychology, National Institute of Occupational and Environmental Health, P.O. Box 3, Raanana 43100


In a previous follow-up study of industrial workers (the CORDIS study, Melamed et al., 1999a) we demonstrated a dose-response relationship between occupational noise exposure levels and all-cause mortality. In that study the type of jobs that workers were engaged in was not taken into account. However, in further analyses of CORDIS data we have found that noise exposure is particularly detrimental to health for workers engaged in complex jobs. Therefore in this 12­year study we attempted to determine the combined effect of job complexity and noise exposure on all-cause mortality in 2606 industrial workers. We divided the workers into four groups based on a combination of either high or low noise exposure, and whether they performed simple or complex jobs. There was an increased risk for all-cause mortality (OR = 1.86, 95% CI = 1.04-3.32), in workers who performed complex jobs under high noise exposure levels compared to those who performed simple jobs under low noise exposure. This remained significant even after adjusting for possible confounding variables. There was a trend for a more pronounced effect among less educated workers, among blue-collar workers, and in those with higher tenure. We conclude that occupational noise exposure is associated with excess mortality risk among workers performing complex jobs.

How to cite this article:
Melamed S, Froom P. The joint effect of industrial noise exposure and job complexity on all-cause mortality - The CORDIS study.Noise Health 2002;4:23-31

How to cite this URL:
Melamed S, Froom P. The joint effect of industrial noise exposure and job complexity on all-cause mortality - The CORDIS study. Noise Health [serial online] 2002 [cited 2021 Oct 17 ];4:23-31
Available from:

Full Text

Evidence arising mostly from cross-sectional studies suggests that chronic exposure to occupational noise may pose a risk to health. The most frequent associations reported were for cardiovascular disease, particularly hypertension. However, this causal relationship is not established because findings have not been consistent (see reviews by Babisch, 1998; Kristensen, 1989; Stansfeld, Haines and Brown, 2000). Occupational noise exposure is related to increased catecholamine secretion, and to changes in the diarnal variation and/or levels of cortisol (Maschke, Rupp and Hecht, 2000; Melamed and Bruchis, 1996). Other extra­auditory effects of noise are sleep disturbances (Stansfeld et al., 2000, Passchier-Vermeer and Passchier, 2000) and possible immunological deficiencies (Passchier-Vermeer and Passchier, 2000) . Therefore, it is biologically plausible that exposure to noise might shorten the worker's life-span.

Almost no study other than our own (Melamed, Kristal-Boneh and Froom, 1999a) explored the possibility of increased mortality risk related to exposure to occupational noise. In that 8-year follow-up study, based on CORDIS data, we reported a dose-response relationship between chronic noise exposure and risk for all-cause mortality. This increased risk was graded and became significant for workers exposed to high noise levels [ > 85dB(A)] for 10 years or more. This association remained significant even after adjusting for age, body mass index, smoking, leisure physical activity and use of hearing protectors (HR=1.96, 95% CI 1.01-3.83). Increased risk was also observed for cardiovascular mortality, but this was not statistically significant, probably because of the small number of workers who died of cardiovascular diseases.

It is quite surprising to find that most studies of the health effects of occupational exposure to noise did not take into account the type of work performed. Evidence obtained from laboratory studies suggests that work characteristics such as task complexity, task difficulty and required effort might modulate the adverse effects of exposure. In a series of experiments, exposure to recorded noise (factory, aircraft or traffic) in combination with mental task performance (binary choice test) resulted in an additional increase in blood pressure, compared to exposure to recorded noise alone (Mosskov and Ettema, 1977a, b). Similar results were obtained in other studies (Carter and Beh, 1989, Ray, Brady, and Emurian, 1984), but not in all (Wu, Huang, Chou and Chang, 1988). The combination of noise and cognitive task performance was reported to be associated with increased muscle tension (Hanson, Schellekens, Veldman, and Mulder, 1993) and increased heart rate, norepinephrine, and cortisol levels (Tafalla and Evans, 1997).

Thus, it is possible that noise exposure might interact with job complexity to affect employee's health. In the CORDIS study we collected data on job complexity as part of the job analysis conducted for the different jobs held by workers in the 21 factories sampled (see 'Methods' section below). In an earlier analysis we used 2­4 years of follow-up data to explore whether the interaction of high noise exposure and noise complexity would effect changes in blood pressure (BP) levels over time. Results showed that among workers exposed to high noise, those with complex jobs showed increases in BP that were more than double shown by those with simple jobs. Under low noise exposure, there was a small increase in BP for workers with complex jobs, but about a three-fold increase in workers with simple jobs. The latter finding suggests that job complexity had a beneficial effect for workers exposed to low noise. The prevalence of elevated BP showed a similar trend. Thus, this study demonstrated for the first time that exposure to occupational noise has a greater negative impact on health among those performing complex jobs (Melamed, Fried and Froom, 2001). The purpose of the present study was to explore if the above results can be extrapolated to the previously observed increased mortality risk in those exposed to high noise levels at work. In other words, will an interaction be found between the complexity of the job and noise exposure levels in predicting all-cause mortality.


Study population. The study participants were 3795 Jewish male workers, aged 20-60 years, who participated in the first phase of the CORDIS (Cardiovascular Occupational Risk Factors Determination in Israel) follow-up study (see Green and Harari, 1995), conducted in 21 industrial plants, 2-4 years apart. A detailed description of the types of plants that were sampled is presented elsewhere (Melamed, Kristal-Boneh and Froom, 1999a). Excluded from this sample were 179 workers under the age of 25 years or those with prior MI or CVA. Of the remaining workers, 1031 had missing values on one or more of the study variables. Thus, the sample was reduced to 2606 workers distributed across 297 work stations. A "Work station" was defined as a group of workers employed in similar physical work and environmental conditions (e.g., control room operators, office workers or workers employed in a given work process). Those who were not included in the final sample were somewhat older (45.5 vs. 43.4 years, respectively, p=0.0001). They had longer tenure (10.9 vs. 9.9 years, p=0.001), less years of education (9.9 vs. 10.4, p=0.001), and a higher percentage were blue-collar workers (83.7% vs. 71.9%, p=0.001). No significant differences between the two groups were observed in body mass index, systolic blood pressure, smoking, and leisure physical activity. The mean age of the workers in the final sample was 43.4±10.9 years, their mean tenure was 9.9±8.1 years and their mean level of education was 10.4±3.15 years. Seventy-two percent were blue-collar and 28% were white-collar workers.


Noise Exposure Levels. Ambient noise levels were measured at each work station where workers were employed. Measurements were conducted by using a Quest sound level meter type SL-215 (area sampling) (Quest Electronics, Oconomovoc, WI, USA), tripod-mounted and adjusted to a height of 150 cm from the floor. Noise levels were sampled twice a day (morning and afternoon) in winter and in summer. Each sampling period lasted one half-hour, during which 5 to 10 readings were taken (depending on noise fluctuations). Results were noted in dB(A) and were averaged for each worker across four sampling periods. The average intercorrelation between noise levels in the 4 sampling periods was >0.90. Noise exposure level was defined by the geometric mean exposure across the four samplings.

Stability of exposure levels over time. The noise exposure level was found to be highly stable; the correlation between noise levels measured twice (2 to 4 years apart) at the same work stations was 0.86 (Melamed et al., 2001). In the present study we used the noise levels measured at phase 1 (1985-1987).

Job Complexity. Job complexity was assessed by averaging an expert's ratings of two items. The first item, "task complexity," provided an overall assessment of the number of elements, decisions, skill level, independence, and sophistication of the employee's job. Rating ranged from 1, representing a very simple job, to 4, representing a very complex job. The second item, "task variety," assessed the diversity of tasks in a given job. Ratings on this item also ranged from 1, representing no diversity, to 4, representing high diversity. These items, which correlated 0.87, were part of the job analysis conducted for the 480 jobs held by the employees in the 21 plants sampled. Other work characteristics included in the job analysis were: type of work (repetitive work, work underload), pay system, type of service/production processes and other general characteristics (e.g., rotation, team work). Job analyses were performed by an experienced rater who participated in the pilot study described below. He observed workers in the same jobs for one day. The reliability of the ratings was evaluated in a pilot study that focused on 48 jobs in two plants from two different industries. Ratings were made by 3 independent raters who observed workers in the same job on two separate days. Inter-rater agreement assessed by Kappa statistic (Cohen, 1977) had a median value of .91.

Inspection of the job complexity scores revealed a bi-modal distribution with two distinct clusters of high and low scores. Based on this finding and given that the study hypotheses were specifically formulated in terms of low and high job complexity, we dichotomized job complexity into low and high on the basis of median split of the score distribution (scores 2-4 and 5-8, respectively).

Mortality data. Follow-up data on mortality was obtained from the National Death Registry (NDR) for the years 1987 through 1998 and all death certificates were re-examined. Further information was obtained from the National Cancer Registry.

We verified the deaths identified by the NDR by contacting, 2690 of the 3795 workers' households during two years following the 8 year follow-up period. The mortality data was reliable since no additional deaths taking place during the 12-year follow-up period were identified; this assured that there was no loss to follow up. We also compared causes of death on the death certificates to that reported in the NDR, and confirmed the cancer deaths using the Cancer Registry. The mortality rate among the workers not included in the study was 9.3% (111/1189) and for those included it was 6.0% (156/2606). There was no difference in the mortality rate between the two groups after adjusting for age.

Confounder variables included age, body mass index (BMI), smoking, leisure physical activity and hearing protector use as reported previously (Melamed et al, 1999a). We added an additional confounder, systolic BP that was found to be significantly related to job complexity and noise exposure in our recent study (Melamed et al., 2001). This enabled us to examine whether the possible association between our predictor variables and mortality was over and above the blood pressure levels.

Data analysis. The significance of the difference between the characteristics of the workers who died and those still alive at follow-up was tested either by applying t tests (for continuous data) or by χ 2 tests (for frequency data). We have used logistic regression analysis to test the hypothesized interactive effect of noise exposure levels [low, high (>80dB(A)] by job complexity levels (low, high) on mortality risk. No such interaction was found, either before or after the inclusion of the possible confounders. Thus, in subsequent regression analysis we tested the possible additive effect of noise and job complexity on the above outcome. This was done by including the 4 combinations of noise exposure levels (low, high) by job complexity (low, high) in the regression, in addition to the confounder variables. Finally, we used stratification analysis to test whether the joint effect of predictor variables on mortality was modified by various variables.


The characteristics of the workers who died and those still alive at follow-up, and univariate analysis results are presented in [Table 1]. This data indicate that among those who died, a higher percentage was employed under high ambient noise. Not much difference was found in this analysis between those in simple jobs versus those in complex jobs. Those who died were significantly older, were smokers, were less engaged in leisure physical activity, had higher blood pressure, were of lower education, and had somewhat higher BMI. No significant difference was found between the groups with regard to hearing protector use.

The risk for all-cause mortality by noise exposure and job complexity levels was estimated through logistic regression analysis. This was done after adjusting for age, body mass index, smoking, leisure physical activity, systolic BP, and hearing protectors use. The results are presented in [Table 2]. By and large the results show the additive effect of noise exposure levels and job complexity on mortality rate. The results indicate that both noise exposure levels and job complexity were associated with increased mortality risk. However, only those workers who performed complex jobs under high ambient noise levels were at significantly higher risk (OR=1.86, 95% CI 1.04-3.32), compared to the reference group of workers performing simple jobs under low noise levels. This was significant even after controlling for several possible confounders.

Finally, we tested interactions between the predictor variables and the confounder variables or other important variables in effecting the outcome studied. This was done through stratification of the variables in question. The results of logistic regressions performed separately for each strata are presented in [Table 3]. The results revealed that the joint effect of the predictor variables on mortality risk was modified by certain variables. This effect was significantly higher among workers with lower education. A similar trend was observed for blue­collar workers, younger workers (aged 10years).


The findings of this study show that the long­term outcome of damage to health resulting from chronic exposure to noise stress is increased mortality risk. They further demonstrate that both noise exposure levels and job complexity independently contribute to mortality risk, but only workers who perform complex jobs under high ambient noise levels are at significantly higher risk. Such workers are at nearly double risk for mortality compared to workers exposed to low noise levels and performing simple jobs, even after controlling for several potent confounders. Furthermore, in our previous study with only 8 years follow-up there were only 95 total deaths, and a significant effect of noise on all-cause mortality was demonstrated only for those exposed to noise levels of 85dB(A) or more. The present study had four additional years of follow-up. More endpoints increased the sensitivity of the study and we were able to demonstrate an increased risk of all-cause mortality in those exposed to 80dB(A) or more. These findings complement and extend our earlier finding from a short-term follow-up study showing that the combination of high noise exposure and performance of complex jobs results in elevated blood pressure levels (Melamed et al., 2001). Taken together, these findings provide almost the first empirical support in a field study to the suggestions that come from laboratory studies. That is, that the combination of performing cognitively demanding tasks under noise distraction is stressful, has physiological and psychological costs (see introduction for review), and may pose a risk to health.

Additional analyses have confirmed the importance of chronicity of exposure. Workers with long job tenure (>10 years) that may serve as a proxy of duration of exposure, were at higher risk for mortality than those with shorter tenure. This result was significant even after adjusting for age. Adjusting for age is necessary because age and tenure are positively correlated. It was important to note that the results were also more significant for workers with lower education levels. Such workers may have higher traditional risk factors but may also defer in psychosocial factors such as job stress, resources to control external demands, or social support (Bucher and Ragland, 1995). They also may be employed more in blue collar jobs and/or work environments that contain additional stressors besides high ambient noise. Indeed, stratification into white and blue collar jobs revealed a similar trend, with blue collar workers having significantly higher mortality risk. For a detailed account of the adverse job and environmental conditions to which blue collar workers are exposed see Melamed et al. (1999b).

Another noteworthy finding was that the association was stronger in younger workers (80dB(A)]. These findings emphasize the importance of considering the type of job performed when determining the risk to health of occupational noise stress.[21]


1Babisch, W. (1998). Epidemiological studies of the cardiovascular effects of occupational noise - a critical appraisal. Noise & Health 1:24-39.
2Bucher, H.C., Ragland, D.R. (1995). Socioeconomic indicators and mortality from coronary heart disease and cancer: A 22-year follow-up of middle-aged men. Am J Pub Health 85:1231-1236.
3Carter, N.l., & Beh, H.C. (1989). The effect of intermittent noise on cardiovascular functioning during vigilance task performance. Psychophysiol. 26:548 -559.
4Cohen, J. (1977). Statistical power analysis for the behavioral sciences. New York: Academic Press.
5Green, M.S. & Harari, G. (1995). A prospective study of the effects of changes in smoking habits on blood count, serum lipids and lipoproteins, body weight and blood pressure in occupationally active men. The Israeli CORDIS study. J. Clin. Epidemiol. 48:1159-1166.
6Hanson, E.K.S., Schellekens, J.M.H., Veldman, J.B.P., & Mulder, L.J.M. (1993). Psychomotor and cardiovascular consequences of mental effort and noise. Hum. Movement Sci. 12:607-626.
7Kristensen, T.S. (1989). Cardiovascular diseases and the work environment: A critical review of the epidemiologic literature on nonchemical factors. Scand. J. Work Environ. Health 5:165 - 179.
8Maschke, C., Rupp, T., Hecht, K. (2000). The influence of stressors on biochemical reactions - a review of present scientific findings with noise. Int. J. Hyg. Environ. Health 203:45-53.
9Melamed, S., Bruhis, S. (1996). The effects of chronic industrial noise exposure on urinary cortisol, fatigue and irritability. A controlled field experiment. J. Occup. Med. 38:252-256.
10Melamed, S., Kristal-Boneh, E. & Froom, P. (1999a). Industrial noise exposure and risk factors for cardiovascular disease: Findings from the CORDIS study. Noise & Health 4:49-56.
11Melamed, S., Yekutieli, D., Froom, P., Kristal-Boneh, E., & Ribak, J. (1999b). Adverse work and environmental conditions predict occupational injuries. The Cardiovascular Occupational Risk Factors Determination in Israel (CORDIS) study. Am.J. Epidemiol. 150:18-26.
12Melamed, S., Fried, Y., Froom, P. (2001). The interactive effect of chronic exposure to noise and job complexity on changes in blood pressure and job satisfaction: A longitudinal study of industrial employees. J. Occup. Health Psychol.6(3): 182-195
13Morrell, S., Taylor,R., Lyle, D. (1997). A review of health effects of aircraft noise. Aust. New Zealand J. Pub.Health 21:221-236.
14Mosskov, J.I., Ettema, J.H. (1977a). III. Extra-auditory effects in short-term exposure to noise from a textile factory. Int. Arch. Occup.Environ. Health 40:174-196.
15Mosskov, J.I., Ettema, J.H. (1977 b). IV. Extra-auditory effects in long-term exposure to aircraft and traffic noise. Int. Arch. Occup. Environ. Health 40:177-184.
16Passchier-Vermeer, W., Passchier, W.F. (2000). Noise exposure and public health. Environ. Health Perspect. 108, Suppl.1:123-131.
17Ray, R.L., Brady, J.V. & Emurian, H.H. (1984). Cardiovascular effects of noise during complex task performance. Int. J. Psychophysiol. 1:335-340.
18Stansfeld, S., Haines, M., Brown, B. (2000). Noise and health in the urban environment. Rev. Environ. Health 15: 43-82.
19Tafalla, R.J., & Evans, G.W. (1997). Noise, physiology, and human performance: The potential role of effort. J. Occup. Health Psychol. 2, 148-155.
20Theorell, T. & Karasek, R.A. (1996). Current issues relating to psychosocial job strain and cardiovascular disease research. J. Occup. Health Psychol. 1:9-26.
21Wu, T.N., Huang, J.T., Chou, P.F., & Chang, P.Y. (1988). Effects of noise exposure and task demands on cardiovascular function. Int.Arch. Occup.Environ. Health, 60, 99-105.