There are conflicting reports as to whether traffic noise is associated with increased risk of stroke. For this reason Stokholm et al. sought to investigate this relationship through a prospective cohort study of Danish workers registries. Specifically, they studied the association of stroke with noise level in a population exposed to high noise level, occupational workers (blue-collar), and a population exposed to low noise level, financial workers (white-collar). They also studied the impact of varying occupational noise level and duration in industrial (blue collar) groups. After adjusting for age, sex, socioeconomic status, calendar year, and employment status and antihypertensives and statins in a sensitivity analysis, their data revealed a 27% increase in stroke in the blue collar, but no association between stoke, level of exposure, duration of exposure.
The strength of this study lies in its large cohort of 116,568 blue-collar and 47,679 white-collar workers, and in assessing different noise levels and duration, analogous to a dose-response assessment. However, presumably due to data availability, they did not adjust for diabetes, atrial fibrillation, coronary artery disease, smoking, and aspirin, each of which maybe independently responsible for stroke development. To adjust for these confounders, the authors could have compared level and duration of noise exposure within the blue-collar group, since they are more likely to share similar baseline characteristics that might negate the impact of confounders. Based on the supplied data, it appears that the reference group included both white and blue-collar groups exposed to low noise levels.
I am not sure why would anyone look for an association between noise level and stroke, as there is no pathophysiological mechanism to explain such a relationship. Nonetheless, should my blue-collar patients ask me about the impact of noise level on their risk of developing a stroke, I can emphatically reassure them based on the lack of dose-response association.