Previous studies of the association between occupational noise exposure and cardiovascular disease (CVD) or risk factors for CVD are primarily either cross-sectional or retrospective, whereas the design of the CORDIS study was both cross-sectional and longitudinal. It had three phases: Phase I was conducted during 1985-87 among 6,016 employees from 21 factories. Recorded were medical, ergonomic, environmental (including noise levels at the various work stations) and psychological data. Phase II was conducted during 1988-90, at 18 of the 21 original factories and included similar data collected from 3,509 subjects. Phase III was conducted during the years 1995-96 and 4,995 workers who participated in Phases I and II completed questionnaires pertaining to medical, occupational and life style variables. Mortality and cancer morbidity data were obtained over an 8 year follow-up period for all subjects.
Results from Phase I, revealed no association between noise exposure and resting blood pressure. Positive association was found for serum lipids in women and in young men. Noise annoyance had an additive effect on this outcome. In addition, recurrent daily noise exposure was found to be associated with elevated acute resting heart rate. Results of Phase II showed that chronic exposure to high noise levels during the 2-4 years of the follow-up resulted in changes of 3.9 mmHg in SBP and 3.3 mmHg in DBP, among workers performing complex jobs. In workers performing simple jobs these changes were 0.3 and 0.4 mmHg. Thus the type of work performed appears to be a significant factor. Results of Phase III revealed that there was a trend for positive association between past noise exposure measured at Phase I and 8 years incidence of cardiovascular morbidity, mortality and total mortality. This trend was statistically significant for total mortality (hazard ratio = 1.97, 95% CI 1.28-4.54) even after controlling for possible confounders. In summary, tests for association between noise exposure and cardiovascular risk factors, or cardiovascular morbidity and mortality, and total mortality have yielded mixed results. Reasons for this are discussed, as well as suggestions for further research.
Keywords: Noise exposure, longitudinal study, cardiovascular disease, risk factors, mortality, blood pressure
|How to cite this article:|
Melamed S, Kristal-Boneh E, Froom P. Industrial noise exposure and risk factors for cardiovascular disease: Findings from the CORDIS Study. Noise Health 1999;1:49-56
|How to cite this URL:|
Melamed S, Kristal-Boneh E, Froom P. Industrial noise exposure and risk factors for cardiovascular disease: Findings from the CORDIS Study. Noise Health [serial online] 1999 [cited 2019 Mar 18];1:49-56. Available from: http://www.noiseandhealth.org/text.asp?1999/1/4/49/31724
| Introduction|| |
Field studies conducted over the past two decades have attempted to demonstrate an association between chronic exposure to occupational noise and cardiovascular disease (CVD) and/or risk factors for CVD. The results obtained were inconsistent. Weak correlation, no correlation and even negative correlations between noise exposure and these endpoints were observed in over fifty percent of these studies (Kristensen, 1989; Sloan, 1991; Smith, 1991 and Babisch, 1998). Moreover, most of the studies were cross-sectional or retrospective in nature, usually with no record of exposure history. No control was made in many studies for other negative work and environmental conditions (such as shift work and exposure to physio-chemical factors) which may coexist with noise exposure and can adversely effect the cardiovascular system. Furthermore, in many studies no direct measure of the noise was available and existence of such exposure was inferred by evidence of hearing loss among the study participants.
In most studies of occupational noise exposure and health, the type of work performed, which theoretically should be an important factor, was not taken into account. There is some suggestive evidence from laboratory studies that noise exposure might interact with task characteristics, such as task complexity, task difficulty and required effort to effect physiological outcomes (Babisch, 1998; Mosskov & Ettema, 1977a and Mosskov & Ettema, 1977b).
Finally, there is evidence that noise annoyed or noise sensitive persons, may be negatively affected by the ambient noise at all levels. This may be manifested in both psychological distress symptoms and sickness absence (Melamed et al 1992 and Melamed et al 1994) or in physiological outcomes (Lercher et al 1993). Thus, it is possible that industrial noise exposure will negatively affect health, particularly in the noise annoyed workers.
The CORDIS study (described below), performed on a large cohort of industrial workers, was designed to address some of the methodological difficulties mentioned above. Taken into account was also the type of work performed by the workers and the personal characteristics.
The CORDIS Study
The CORDIS study (Cardiovascular Occupational Risk Factors Detection in Israel) is a multidisciplinary, cross-sectional and longitudinal epidemiological investigation aimed at identifying occupational risk factors for cardiovascular disease. Data collected from each participant at the worksite, covered a multitude of medical, biochemical, ergonomic, environmental (including noise exposure levels at the various work stations) and psychological variables see [Table - 1]. Workers were examined in the non-fasting state between 7:00 A.M. and 4:00 P.M. in a quiet air-conditioned room. Both medical and psychological variables were measured once, and this may have been at different times of day for different workers. The time of day was noted. Participation was on a volunteer basis, free of charge and on company time. The CORDIS study was conducted in three phases: Phase I of the study was conducted during 1985-87 among 6,016 employees (4,318 males and 1,698 females) at 21 industrial plants. The plants represented six different sectors: (1) metal work, (2) textile, (3) light industry, (4) electronics, (5) food and (6) plywood. The response rate was over 60%, and failure to participate was largely due to technical and logistic reasons rather than to workers' refusal. Almost 75% of the subjects were blue-collar workers. Phase II of the study was conducted during 1988-90, at 18 of the 21 original plants. The total number of subjects examined in this phase was (2,235 males and 1,274 females) of these 1,347 males and 598 females had also participated in Phase I. The data collected was similar to that in Phase I. Phase III was conducted during the years 1995-96. We were able to identify 4,995 workers (3,448 males and 1,547 females) who participated in Phase I and II, and asked them to complete a self-report questionnaire pertaining to medical, occupational, life style and perceived health variables. We also obtained mortality data for the years 1987 - 1994 from the National Death Registry (NDR) of the Ministry of Interior, and cancer morbidity data from the National Cancer Registry.
| Noise Measurements|| |
Ambient noise levels at each work station were measured with a Quest sound-level meter; sampled once in the morning and once in the afternoon during the winter and summer. Each measurement was conducted for a half-hour period during which 5-15 readings were taken. Results were noted in dB(A) and were averaged for the four measurement periods. (For more details, see Kristal-Boneh et al 1995)
Noise exposure levels were measured during one day in summer and winter, using a Quest M-27 noise-logging dosimeter. Results were noted in Leq and were averaged for the two sampling periods.
| Results|| |
Phase I: Cross-sectional study
Data collected in this phase was used to examine the association between noise exposure and CVD risk factors while controlling for several possible confounders: age. BMI, physical effort, shift work, work monotony, smoking, alcohol consumption and leisure physical activity. Summarised below are the major findings.
Effect of noise exposure on blood pressure and heart rate
This effect was studied in 3,105 blue-collar male and female workers (Kristal-Boneh et al 1995). Heart rate (HR) and blood pressure (BP) were measured in different workers at various times during the workday. By and large, univariate results indicated that noise exposure levels correlated positively with resting HR in males (r=.08) and (r=.06) in females. In addition, a significant and positive correlation was observed with diastolic blood pressure (DBP) for females (r=.08) but not for males. No correlation was observed with systolic blood pressure (SBP) for both sexes. A subsequent multivariate analyses which included also the time of day and other possible confounders, revealed no association between noise exposure and blood pressure. Resting HR, however, was found to be associated with noise intensity but this was related to time of day. We plotted the mean resting heart rate against time of day for workers exposed to high [>80 dB(A)] and low noise, and no difference was evident with respect to morning heart rate in either sex. After 4 hours of noise exposure for males (and less time for females), individuals who were exposed to high noise had higher heart rates; however, in females this was not observed at the end of the workday. We concluded that the recurrent daily exposure to high noise at work has an acute effect on resting HR. This was replicated and extended the findings of laboratory studies. On the other hand, no chronic effect of noise was found for both resting HR and blood pressure.
Effect of noise exposure and noise annoyance on serum lipid levels
Blue-collar workers were examined again to test yet another possible pathway that links noise exposure to CVD, namely, elevated serum lipid levels (Melamed et al 1997). We also included in this examination noise annoyance as a substitute for individual sensitivity to noise. We found that young men (i.e., < 44 y of age) exposed to high noise levels (> 80 dB[A]) had higher total levels of cholesterol (p=.023) and triglycerides (p=.001), as well as high cholesterol ratio (p=.038), than young men exposed to low noise levels, even after controlling for confounding variables. In women, or in older (> 45 y) men, noise did not affect serum lipid/lipoprotein levels.
Noise annoyance covaried independently with total cholesterol (p=.022) and high-density lipoprotein (p=.0039) levels in young men and with total cholesterol (p=.035), triglyceride (p=.035), and high-density lipoprotein levels in women (under high noise exposure conditions) (p=.048). Noise annoyance and noise exposure levels had an additive effect on cholesterol levels. Young men who scored high on both variables had a 15-mg/dl higher mean cholesterol level (95% confidence interval [CI]=7.2, 22.8; p=.0003) than those who scored low on both variables; in women, the corresponding difference was 23 mg/dl (95% CI=1.5, 42.9; p=.019).
Phase II: Short-term follow-up
Chronic exposure to noise, job complexity and blood pressure evaluation with time
As indicated earlier in the literature review, most of the studies to date were cross-sectional or retrospective in nature. The two-four years of follow-up (Phase II of the CORDIS study) provided us with an opportunity to examine the impact of objectively measured chronic exposure to industrial noise. We were able to identify 653 workers, who continued to be employed at the same work stations throughout the follow-up period, and 518 workers who changed their work stations during this time. The correlation between noise exposure levels at Time 1 and Time II for workers who remained in the same work station was 0.86. For workers who had changed work stations, this correlation was 0.48.
One of the parameters recorded during Phase I of the study was job complexity. Job complexity (high/low) was classified on the basis of job analysis, of the 480 jobs held by the workers in the 21 factories sampled. Job complexity was assessed by two items: (1) Task complexity - provided an overall assessment of the number of elements, decision making, skill levels, independence and sophistication of the worker's job. (1 - a very simple job, 4 - a very complex job). (2) Task variety - assessed the diversity of tasks in a given job (1 - no diversity, 4 - much diversity). These items correlated 0.87. Having data on job complexity enabled us to test for the first time outside the laboratory, a possible interaction between noise exposure and job complexity in affecting increase in BP with time. Thus, we examined the effect of high noise exposure (> 80 dB (A)) on BP change (Time 2 - Time 1 BP levels) among workers performing simple and complex jobs. The preliminary findings from ANCOVA tests applied to the data, showed that significant results were obtained for workers who remained in the same work stations during the follow-up period. Changes of 3.9 mmHg in SBP (p<.01) and 3.3 mmHg in DBP (p<.01) were observed among workers performing complex jobs. The corresponding changes for workers performing simple jobs were 0.3 mmHg (n.s) and 0.4 mmHg (n.s). This finding was obtained even after controlling for several possible confounders: Time 1 BP levels, age, BMI, sex, family history of hypertension, tenure and of hearing protectors use. No significant results were obtained for workers who changed their work stations during this time.
Thus, these findings demonstrated (a) that BP changes with time occurred only in workers who continued to be exposed to high ambient noise throughout the follow-up period, and (b) that occupational noise exposure may lead to BP elevation only in workers performing complex jobs.
Phase III: Long-term follow-up
Noise exposure, cardiovascular morbidity and mortality, as well as total mortality
The data collected during Phase III of the CORDIS study pertained to morbidity and mortality of the participants in Phase I of the study during 8 years. This data was used to examine if noise exposure levels in Phase I are predictive of morbidity and mortality rate during the follow-up period.
The analyses performed so far, concerned 2,697 Jewish males for whom we have complete data. Ninety-five of them died during the follow-up period, of whom 37 died from cardiovascular diseases (CVD). In addition, there were 89 new cases of workers with CVD. We compared these endpoints across three groups of workers: (1) those who were exposed to high noise level (> 85 dB (A)) at time 1 and had tenure of > 10 years, and (2) those exposed to ambient noise levels between 75-84 dB(A), and (3) a referent group of workers exposed to noise level <75 dB(A). The extreme group of workers exposed to high noise levels for an extended period of time (group 1) was chosen because, theoretically, they were expected to differ maximally from other groups with respect to the endpoints studied. However, in separate analyses both noise exposure levels and tenure were entered as continuous variables. Hazard ratios were computed using Cox's proportional hazard models.
The crude hazard ratio (HR) for CVD mortality for the extremely exposed group, compared to the non-exposed group, was 1.25 (95% CI 0.453.49) see [Table - 2]. The HR was 1.77 (95% CI 0.62-5.03) after adjusting for age, body mass index, smoking, leisure physical activity and use of hearing protectors. In order to avoid overcontrolling (see Babisch, 1998) no adjustment was made for hypertension or hyperlipidemia. The respective HRs for the moderately exposed workers were 1.15 (0.572.32) and 1.28 (0.63-2.54). Thus, there is also a trend for increased CVD mortality with increased post exposure to noise, but this was not statistically significant. A similar finding was obtained when CVD morbidity and mortality was combined (results are not shown). On the other hand, significant results were obtained for total mortality see [Table - 3]. The crude HR for the moderately exposed group was 1.70 (1.082.65), but this became non-significant after adjusting for possible confounders [HR = 1.54 (0.97-2.46)]. However, for the extremely exposed group the crude HR was 2.41 (1.284.54) and it remained significant after adjusting for possible confounders [HR = 1.97 (1.01-3.83).
No interactive effect of high noise exposure and tenure on risk for CVD mortality, CVD morbidity and mortality, or total morbidity was found. When noise exposure levels and tenure were entered as continuous variables, they did not significantly add to the model.
Taken as a whole, these findings showed that there was a trend for positive association between past exposure to high ambient noise and cardiovascular morbidity, and mortality. Such a trend, however, was significant for total mortality. It remained significant even after adjustment for possible confounders. A major shortcoming of this follow-up study, was that no information was available on noise exposure in the three groups after Phase I.
Discussion and implication for further research
Integration of the findings across all three phases of the CORDIS study highlight the complexity of studying the health effects of noise, and are helpful in generating suggestions for further research.
The implicit assumption in cross-sectional studies is that current levels of noise exposure, reflect past exposure. Data from Phase II of our study indicate that about 45% of the workers changed their work station during the 2-4 years of follow-up. By doing so, their noise exposure levels increased, decreased or remained unchanged. All had accumulated the same year of tenure at the factory. This finding, by implication, may explain the absence of association between high noise exposure and resting BP in the cross-sectional data (Phase I). Workers currently exposed to high noise levels may have been exposed to low noise levels in the past, and vice versa. Thus, no chronic exposure to noise could be ascertained.
It was interesting to note in the same data, that recurrent exposure to noise has an acute effect on HR elevation. This by itself may constitute a risk to the cardiovascular system. Studies have shown for example, that those displaying HR reactivity to stress may have an accelerated progress of atherosclerosis (Cinciripini, 1986). Thus, it is important to follow-up workers who display a consistent HR response to daily noise exposure at work.
Results from Phase I also showed positive association between noise exposure and serum lipid levels. Increased plasma lipids levels are associated with current stress (Niaura et al 1992). Hence the reason for the difference from the negative results obtained with BP. This important finding points to an alternative pathway by which noise exposure might be associated with CVD. Another important finding was that noise annoyed workers showed even higher lipid levels, compared to non-annoyed workers, when exposed to different noise levels. Findings from the CORDIS study suggests that noise annoyance workers were distressed by noise even at low exposure levels (Melamed et al 1994). Thus, further studies are needed to examine whether chronic exposure to noise results in prolonged elevation of plasma lipids, particularly among the noise annoyed workers.
The preliminary findings from Phase II of the study provides new evidence and insights for further research. Chronic exposure to high noise level, objectively verified, was found to be associated with a significant increase in both SBP and DBP over a period of 2-4 years. This finding, if replicated, will point to the need for additional prospective studies in which noise exposure is periodically monitored. Furthermore, these findings revealed that BP elevation was observed only among workers performing complex jobs. This, in future prospective studies one should also consider the type of jobs performed.
The novel finding here of long term effect on total mortality of past exposure to high industrial noise levels (>85 dB(A)) for 10 years or more, is of high significance. It suggests that noise may adversely affect health in general and this becomes apparent after a sufficient period of follow-up. No significant long term effect on cardiovascular morbidity or mortality was uncovered. This may have at least two alternative interpretations: (1) Because of the small number of cases, the 95% CI is wide. With further follow-up this outcome may become significant. (2) Past long term exposure to high ambient noise may affect other body systems beyond the cardiovascular system, and this has led to increased premature mortality.
Recommendations: Based on findings obtained in the CORDIS study, and other studies in this area, we suggest the 3 points where further studies needed to provide more definitive answers concerning the possible impact of industrial noise on health.
a) Longitudinal studies are needed in which noise exposure is monitored periodically. A number of work and environmental conditions, such as shift work and exposure to physiochemical agents should also be recorded and controlled as possible confounders.
b) The characteristic of the work performed should also be taken into account, in particular, job complexity or job difficulty. The type of job performed may interact with ambient noise exposure to effect physiological outcomes and health endpoints.
c) Attention should also be paid to the noise annoyed or noise sensitive workers. Such personal characteristics may have an additive effect to that of noise on workers' outcomes.
d) Further studies should explore possible adverse effects of noise on other body systems beyond the CV system.
| References|| |
|1.||Babisch, W. (1998) Epidemiological studies of the cardiovascular effects of occupational noise - A critical appraisal. Noise and Health, 1: 24-39. |
|2.||Cinciripini, P.M. (1986) Cognitive stress and cardiovascular reactivity II. Relationship to atherosclerosis, arrhytmius, and cognitive control. Am Heart J. 112: 1051-1065. |
|3.||Kristal-Boneh, E., Melamed, S., Harari, G., Green, M.S. (1995) Acute and chronic effects of noise exposure on blood pressure and heart rate among industrial employees: The CORDIS study. Arch Environ Health. 50: 298-304. |
|4.||Kristensen, T.S. (1989) Cardiovascular diseases and the work environment: A critical review of the epidemiologic literature on non-chemical factors. Scand J Work Environ Health. 5: 165-179. |
|5.||Lercher, P., Hortnagle J., Kofler, W.W. (1993) Work noise annoyance and blood pressure: Combined effects with stressful working conditions. Int Arch Occup Environ Health. 65: 23-28. |
|6.||Melamed, S., Luz, J., Green, M.S. (1992) Noise exposure, noise annoyance and their relation to psychological distress, accident and sickness absence among blue-collar workers. The CORDIS study. Isr J Med Sci. 28: 629-635. |
|7.||Melamed, S., Rabinowitz, S., Green, M.S. (1994) Noise exposure, noise annoyance, use of hearing protection devices and distress among blue-collar workers. Scand J Work Environ Health. 20: 294-300. |
|8.||Melamed, S., Froom, P., Kristal-Boneh, E., Gofer, D., Ribak, J. (1997) Industrial noise exposure, noise annoyance and serum lipid levels in blue-collar workers: The CORDIS study. Arch Environ Health. 52: 292-298. |
|9.||Mosskov, J., Ettema, I. (1977a) Extra-auditory effects in short-term exposure to noise from a textile factory. Int Arch Occup Environ Health. 40: 174-196. |
|10.||Mosskov, J., Ettema, I. (1977b) Extra-auditory effects in short-term exposure to aircraft and traffic noise. Int Arch Occup Environ Health. 40: 177-184. |
|11.||Niaura, R., Stoney, C.M., Herbert, P.M. (1992) Lipids in psychological research: The last decade. Biol Psychol. 34: 1-43. |
|12.||Sloan, R.P. (1991) Cardiovascular effects of noise. In Fay, T.H. (ed). Noise and Health. New York, The New York Academy of Medicine. Pp. 15-26. |
|13.||Smith, A. (1991) A review of the non-auditory effects of noise and health. Work and Stress. 5: 49-62. |
|14.||Tafalla, R.J. Evans, G.W. (1997) Noise, physiology, and human performance: The potential role of effort. J Occ Health Psychol. 2: 148-155. |
Department of Occupational Health Psychology, Occupational Health Institute, Ra'anana
Source of Support: None, Conflict of Interest: None
[Table - 1], [Table - 2], [Table - 3]