Occupational exposure to noise and dust in Swedish soft paper mills and mortality from ischemic heart disease and ischemic stroke: a cohort study

An important finding in this large longitudinal cohort study in soft paper workers was that occupational exposure to noise over 90 dBA, as well as exposure to noise ≥ 85 dBA, was associated with increased mortality from ischemic heart disease. There was also an increased myocardial infarction mortality in workers with noise exposure > 90 dBA, but there were increased myocardial infarction mortality in both higher and lower occupational noise levels, which weakens the support of a causal association. The study indicated that the combination of high noise exposure and high paper dust exposure may further increase the mortality, especially regarding ischemic stroke. However, the influence of uncontrolled bias like tobacco smoking is probably of importance.

The finding in the present study addresses a key knowledge gap regarding whether occupational noise exposure increases cardiovascular mortality (Teixera et al. 2021). Previous studies have mostly evaluated the broader outcomes of cardiovascular disease or ischemic heart disease. In addition to ischemic heart disease, we have used the more specific outcome of myocardial infarction. Ischemic heart disease includes, in addition to myocardial infarction, angina pectoris, complications to myocardial infarction and chronic ischemic heart disease. The ischemic heart disease mortality we observed was in general lower compared to myocardial infarction mortality. In the cited review by (Teixera et al. 2021), four cohort studies were identified where ischemic heart mortality was analyzed in relation to occupational noise exposure (Teixera et al. 2021; Gopinath et al. 2011; Suadicin et al. 2012; Pettersson et al. 2020). The pooled effect-estimate in that review indicated only a very small risk increase (RR = 1.03, 95% CI 0.93–1.14) with the range from 0.97 to 1.44 (Teixera et al. 2021). However, this comparison should be judged with caution as the background populations may differ. There was no attempt in that review to separate papers studying myocardial infarction, an outcome that needs further assessment (Troke et al. 2021). Hence, we consider that our study shows an increased ischemic heart mortality among workers occupationally exposed to noise levels exceeding 90 dBA. When we stratified according to gender, we clearly observed increased mortality among men in the high occupational noise group. However, among women the observations were less consistent with increased mortality both in the low and high occupational noise groups. Of note, the power was also low among the women. This may reflect influence of uncontrolled bias, like tobacco smoking.

Regarding cerebrovascular mortality and ischemic stroke mortality, the patterns are less clear.

In published papers, stroke is often defined as ICD10 I60 – I69, or similar codes in the earlier classifications (Gopinath et al. 2011; Pettersson et al. 2020). This is a broad definition that also includes subarachnoid hemorrhage, I60, and transient ischemic attacks, I65. Gopinath et al. did not find any association between occupational noise exposure exceeding five years and mortality due to stroke (Gopinath et al. 2011). Stroke was defined in that study as ICD 10 I60–I69, and the hazard ratio was 1.01, 95% CI 0.62–1.66. In a Swedish study, also with a broad definition of stroke (I60–I69), workers exposed > 85 dBA had an increased stroke mortality, 1.19, 95% CI 1.03–1.38 (Pettersson et al. 2021). For stroke, we have used a more specific definition (I63–I64), ischemic stroke. We did not observe any clearly increased ischemic stroke mortality or cerebrovascular mortality in relation to different groups of occupational noise exposure, even if there was a signal of increased ischemic stroke mortality in workers with > 90 dBA exposure (SMR 1.48, 95% CI 0.99–2.21). Of note, we observed increased ischemic stroke mortality among the subgroup with high occupational noise exposure and high exposure to soft paper dust (SMR 1.83, 95% CI 1.12–2.98). It has been argued that occupational exposure to noise exceeding 80 dBA probably will not increase the risk for stroke (Kolstad et al. 2013; Stokholm et al. 2013). Our results do not provide support for a causal relationship between occupational noise exposure and stroke, as residual confounding such as smoking may be present.

In our main analyses we classified occupational noise into three categories, < 85 dBA, 85–90 dBA and > 90 dBA, as previously recommended (Teixera et al. 2021). That allowed us to look into internal comparisons such as exposure–response relationships. We observed increasing mortality with increasing noise exposure, which may indicate a relationship between occupational noise exposure and all-cause mortality, ischemic heart disease and ischemic stroke. However, the study comprises a long time period when occupational exposures, lifestyle habits including smoking habits, and mortality due to cardiovascular diseases had changed a lot, we applied comparisons with calendar-year specific national mortality rates. Consequently, we did not perform internal comparisons. A variety of mechanisms have been proposed to explain the relationship between noise exposure and cardiovascular diseases. Health effects of noise exposure can be mediated through stimulation of the central stress response, which can result in disturbed sleep, decreased heart rate variability and increase stress hormone levels. Taken together, these changes can directly facilitate the development of cardiovascular disease (Zaman et al. 2022).

Regarding the stratified analyses, the combinations of high noise and high paper dust exposure resulted in a further increased all-cause mortality and further increased mortality from ischemic heart disease, myocardial infarction but also from ischemic stroke. In our previous study we observed among workers with high exposure to soft paper dust an increased cerebrovascular disease mortality (SMR 1.30, 95% CI 1.00–1.67), but no clear increase in all-cause mortality (SMR 1.08, 95% CI 0.99–1.17) or ischemic heart disease (SMR 1.06, 95% CI 0.90–1.26) (Torén et al. 2020). One possibility is that the findings in the present study are the results of an interaction between high noise exposure and dust, as exposure to dust can induce inflammation in the airways, which represents an additional mechanism for cardiovascular disease (Joshi et al. 2022). However, it can also be explained by different smoking habits. In one of the mills we had information about smoking habits, and workers with high exposure to soft paper dust were more often current smokers as compared to low-exposed workers (30% vs 22%). Hence, we consider a possible explanation to the higher mortality, especially from myocardial infarction, in the subgroups with high paper dust exposure could be explained by residual confounding due to tobacco smoking.

Our study has several strengths. Most importantly, we did not rely on self-reported data, but rather used objective data. The exposure assessments were based on validated job-exposure matrices for noise and dust (Neitzel et al. 2016, 2020). The outcomes were also based on causes of death from mortality registers, not on self-reported disease. Another strength was the prospective design with cohort inclusion based on personnel files from the mills which resulted in a cohort that comprised almost all workers that have been employed in these mills, and gave us a comprehensive perspective of soft paper industry employment and exposures. Furthermore, due to our use of personal identification numbers, our medical follow-up had a very high degree of completeness. We also evaluated mortality estimates for men and women separately. A final strength is our simultaneous consideration of paper dust (a chemical hazard) and noise (a physical hazard).

The study has a number of limitations. One limitation is the healthy-worker survivor effect (McMichael 1976). Due to noisy, dusty, and demanding work in paper mills, less healthy workers tend to terminate their employment or be transferred to lower exposed positions in the mills. This can result in weakened associations between noise exposure and outcomes. This bias is likely further accentuated by our use of the total Swedish population as the comparison group, rather than the employed fraction of the population. On the other hand, only newly-employed workers were included in the cohort, which probably diminished healthy-worker survivor effect. Shift work is also a possible confounder, as paper mill workers often have shift work and shift work is clearly linked to increased risk of cardiovascular disease (Torquati et al. 2018). Another important limitation, as mentioned above, is the lack of information on smoking habits for the entire cohort. Since myocardial infarction, ischemic heart disease, and ischemic stroke are associated with smoking, residual confounding was present in our analyses. Another limitation is the existing errors in the applied JEMs, especially regarding the dust JEM for the period before 1970, when exposures were assigned via expert consensus, as no measurement data were available.

In conclusion, we found that occupational noise levels > 90 dBA was associated with increased mortality from ischemic heart disease, but the evidence for an association with myocardial infarction is questionable. Our results do not provide support for a causal relationship between occupational noise exposure and stroke. Workers with joint exposure of high noise levels and high paper dust exposure seemed to have further increased mortality, but that may be due to uncontrolled confounding of tobacco smoking. We have observed similar results for occupational noise exposure ≥ 85 dBA. The results underscore that occupational noise levels must be reduced, and that workers need to be protected to prevent noise-induced hearing loss and to reduce the risk for non-auditory health effects.

留言 (0)

沒有登入
gif