A retrospective cohort study of Parkinson's disease in Korean shipbuilders

We performed a retrospective cohort study in South Korea to clarify the role of occupational exposure, especially to welding, in the etiology of Parkinson's disease (PD).

We constructed a database of subjects classified into an exposure group (blue-collar workers) and a non-exposure group (white-collar workers) in two shipbuilding companies. Jobs of blue-collar workers were categorized into the first group of welding, the second group of fitting, grinding and finishing, cutting, and the other group. To determine new cases of PD during the follow-up period (1992-2003), we used the physician billing claims database of the National Health Insurance Corporation. For the detected PD patients in the physician billing claims database, a neurologist in our research team confirmed the appropriateness of each diagnosis by reviewing medical charts. Based on the review, we confirmed the numbers of new cases of PD and calculated the relative risk (RR) and the 95% confidence intervals (CI) by Cox regression analysis.

In a backward selection procedure, 'age' was a significant independent variable but exposure was not. Furthermore, the RR in welders (high exposure group) was also insignificant and less than that in others (very low exposure group).

This longitudinal study of shipbuilding workers supports our previous case-control studies suggesting that exposure to manganese does not increase the risk of PD.