There are several important findings of this study. First, the prevalence of parkinsonism we found in these welding-exposed active workers was quite high considering that the frequency of workers with UPDRS3≥15 approached the frequency of parkinsonian signs found in the elderly.[23
] These subjects had not been diagnosed with any neurologic condition but the neurologic findings do appear to be associated with functional impairments. For this study, we used more conservative criteria than in a previous study and still found a prevalence of parkinsonism of 12%, as compared to 6%–12.8% in welders in Alabama.[6
] Of course, with only half of eligible workers examined, thus far, the prevalence may be somewhat lower if the majority of the remaining workers had normal examinations. Nevertheless, the prevalence of parkinsonism in active workers is still quite high. These findings are not surprising given substantial evidence of clinical and subclinical motor abnormalities associated with occupational Mn exposure. A study of workers in a Mn oxide and salt producing plant found that those exposed to low levels of Mn (approximately 1mg/m3) had slower simple reaction times on a standardized reaction time test, and more hand tremor, as measured by a standardized hand steadiness assessment.[24
] Similarly, foundry workers in Sweden exposed to even lower Mn levels demonstrated slower reaction time, reduced finger-tapping speed, reduced tapping endurance, and diadochokinesis.[25
] Manganese exposed workers in a Mn alloy facility had slower computerized finger tapping scores and less hand steadiness in another study.[26
] None of these studies investigated the impact of parkinsonian findings on worker health status. It is important to note that our study does not specifically address the cause-effect relationship between welding fume and these parkinsonian findings. Detailed cumulative lifetime dose reconstruction is ongoing and will be used to determine the relationship between welding dose exposure and parkinsonism.
A second important finding of this study is that welding-exposed workers with parkinsonian findings have substantial reductions in health status affecting a broad range of categories. The QoL impairments manifested in this population extend far beyond what would be expected in a healthy population of workers and may have important occupational and social implications for this working population. Several studies have shown that workers with depressive symptoms, such as poor emotional well-being, are at increased risk of injury in the workplace.[27
] There is also an increased risk of injury in workers with at least one disease/illness as compared to healthy workers.[28
] While the relationship between parkinsonism and perceived health status/QoL and job performance in welders needs further study, it is possible that these workers may experience decreased productivity and/or may be at risk for additional occupational injuries and associated work absenteeism.
The elevations in PDQ39 total and subscores in parkinsonian welders in our study were within the range of typical PDQ39 scores for early PD patients.[29
] Previous studies in patients with Hoehn & Yahr stages I and II PD, and the data from patients in our movement disorders center, demonstrate that PD patients present with PDQ39 scores in the teens to twenties. In our welders, the most significant elevations in PDQ39 subscores were in ADL's, communication, and stigma. While the communication and ADL’s abnormalities probably reflect the well described motor impairments associated with Mn exposure, the emotional well being abnormalities may be consistent with the cognitive behavioral effects seen in Mn exposed workers.[30
] Emotional well being, communication, and cognition subscores of the PDQ39 were higher in our welders with parkinsonism than our early PD patients. More detailed study of the neuropsychological impacts of chronic welding fume exposure in healthy workers is clearly warranted, but was beyond the scope of this study which focused on the impact of parkinsonism in welding exposed workers on health status.
Although our study demonstrates a relationship between abnormalities in health status and parkinsonism in welding exposed workers, the number of symptoms subjects experienced was quite modest. This may reflect the relatively lower sensitivity of traditional PD symptoms for detecting parkinsonism in this population compared to the more detailed QoL health status questions contained in the PDQ39. Of the individual symptoms, micrographia was most predictive of having parkinsonism, but the modest number of subjects with parkinsonism reduced the precision of these estimates. Frequency of positive answers on the symptom questionnaire was not particularly predictive of parkinsonism. However, if cognitive impairments are the first evidence of neurotoxicity in welding exposed workers, symptom questionnaires focused on motor symptoms may not be sufficient to identify cases reliably. Ideally, a combination of PD symptoms and health status questionnaires could be used to identify workers who may have developed parkinsonism and could be used to screen workers as part of routine occupational health examination. This would be far more practical than expert physical examinations in all workers and will be an important long term goal of this study.
There are several potential limitations to this study. Although this is a reasonably large, worksite-based sample of welding exposed workers, assessing a larger population may alter the relationship between parkinsonism and the measures of perceived health status found in this study if there is a volunteer bias among symptomatic welders. In addition, the PDQ39 may not be as specific for PD as suggested in the literature. None of the PD specific QoL or health status questionnaires have been adequately validated for their specificity to PD, although they do have face validity for the known health effects of PD. However, this population of active workers was quite healthy, based upon medical history, so we believe that co-morbid disease does not likely explain these findings. We were careful to exclude subjects with any co-morbid neurologic disease that would affect the motor ratings. Similarly, the UPDRS was designed to monitor PD subject progression in clinical trials so it is possible that nonspecific motor findings may explain some of the neurologic findings. Accordingly, to improve diagnostic validity, we established a threshold for the diagnosis of parkinsonism, UPDRS>15, which should minimize the likelihood that cases would consist of subjects with nonspecific motor findings. This study was therefore an attempt to test the assertion that health status impairments found in parkinsonian subjects among welders were similar to those typically seem in idiopathic PD. Nevertheless, we cannot exclude the possibility that nonspecific motor signs may contribute, in part, to the health status impairments. Finally, as previously noted, determining the dose-response relationship between welding fumes and parkinsonism was beyond the scope of this report, but with the development of validated metrics for cumulative welding exposure, we expect to investigate the relationship between lifetime welding exposure and perceived health status as well as clinical disease definitions.
In summary, workers exposed to welding fumes have previously undiagnosed motor abnormalities that are associated with measurable impairments in health status and QoL, within the range seen in early idiopathic PD. This study adds to previous literature, suggesting that welding exposed workers frequently demonstrate clinical signs of PD, by investigating the health status and QoL impact of these findings. The impact of these abnormalities on work performance and safety will be an important focus of future research.