Welders exposed to manganese-laden fumes develop a form of parkinsonism, but do these symptoms signify a progressive, degenerative disorder? The answer may be yes, according to a new study published in Neurology on December 28. Researchers led by Brad Racette at Washington University in St. Louis reported that the more workers were exposed to welding fumes, the more their parkinsonian symptoms worsened. While it is unclear how manganese-triggered parkinsonism relates to typical PD, the progressive symptoms in welders point to neurodegeneration, Racette said.

Neurotoxicity caused by exposure to manganese—a disorder known as manganism—has been documented since the mid-1800s, when miners inhaled particulates of the heavy metal (see Couper, 1867, and Rodier, 1955). Working conditions have since improved, and now welders exposed to fumes emanating from manganese rods bear the brunt of exposure, albeit at far lower levels than miners of decades past. Brain imaging studies conducted by Racette and colleagues pointed to an accumulation of heavy metals in the basal ganglia and reduced dopaminergic signaling in the caudate nuclei of welders, and also revealed that more than 15 percent of welders exhibit symptoms of parkinsonism (see Criswell et al., 2012; Apr 2011 news; and Racette et al., 2012). 

For the current study, Racette wanted to determine whether those symptoms worsened with time. Drawing on union membership lists, the researchers recruited 886 welders and co-workers exposed to welding fumes at two shipyards and one heavy machinery fabrication shop in the midwestern United States. At baseline and follow-up visits, Racette, or co-author Susan Criswell, assessed each participant’s performance on the Unified Parkinson Disease Rating Scale motor subsection part 3 (UPDRS3). The 108-point test measures a range of parkinsonian symptoms, including limb bradykinesia, rigidity, and tremor. Racette told Alzforum that few people ever reach the top of the scale—those in the advanced stages of the disease score around 60 points, while newly diagnosed PD patients tend to score between 15 and 20. At baseline, more than 15 percent of the workers scored at least 15 points.

Of those enrolled in the study, 398 made it to at least one follow-up visit. The researchers followed these participants for an average of 4.2 years, with a maximum follow-up of a decade. On average, UPDRS3 scores increased by 0.32 points annually across the entire cohort. Welders’ scores increased by 0.57 points per year, compared to 0.45 points for welder helpers, and 0.16 points for non-welding workers. By contrast, workers who were not exposed to welding fumes experienced no worsening of parkinsonian symptoms.

The researchers estimated cumulative manganese exposure in the workers by considering daily exposure levels and years on the job. This indirect measure of exposure correlated with the progression of parkinsonian symptoms, with each additional milligram Mn/m3 per year at baseline bumping up the annual change in UPDRS3 score by 0.24 points. The association was primarily driven by workers who performed flux core arc welding in confined spaces, as this type of welding generates the highest amount of particulates. These correlations largely held up even when accounting for age, sex, race, tobacco smoking, alcohol consumption, and pesticide exposure.

The study tested only active workers, not those who had quit or retired. The researchers acknowledged that this could cause a “healthy survivor effect,” in which the study cohort—especially those on the job the longest—would be enriched for workers most resistant to parkinsonism. In support of this idea, the researchers found that the link between manganese exposure and progression was strongest in workers who had been exposed for less than five years. 

Do the findings suggest that parkinsonism caused by manganese exposure is a neurodegenerative disease, like PD? Racette said it is difficult to answer that question. Compared to typical PD patients, people with this form of parkinsonism are affected on both sides of the body, tend to have less tremor, and respond more poorly to dopaminergic therapy. “It’s a slow, stiff version of Parkinson’s,” Racette said.

According to Jeffrey Kordower of Rush University in Chicago, these and other differences between PD and manganese-parkinsonism indicate that these are two entirely different syndromes. Kordower, who consults for the manganese industry, added that while the two disorders may be different, the new findings suggest that parkinsonian symptoms associated with manganism can indeed worsen with continued exposure.

To further characterize this form of parkinsonism, Racette is performing longitudinal structural MRI and dopamine-PET imaging on a cohort of welders to look for signs of neurodegeneration or waning dopaminergic function. He has not yet secured funding to monitor progression in workers after they quit or retire. He said he hopes these and other studies will motivate employers to beef up protection efforts for their workers. Current safety standards are not enforced, and fall well above the levels of exposure experienced by welders.—Jessica Shugart


  1. The instrument used to determine the rate of change over the follow-up period, the UPDRS3, rates motor signs on a scale of 0 to 4, with 0 being absent and 4 being very severe. Because individuals with absent or very mild signs at baseline might be expected to show little or no progression over the follow-up period, i.e., to remain non-symptomatic, those groups with higher percentages of such members would be expected to show the lowest rates of progression. Conversely, those groups with more members with greater initial levels of impairment might be expected to show higher rates of progression. The findings therefore could primarily reflect the initial level of impairment of each of the groups at baseline.  

    To determine the role of the baseline scores on the UPDRS3 on the results, the baseline score on this instrument should have been controlled or the findings stratified by initial scores on the UPDRS3. Using non-linear outcomes like the UPDRS3 score to assess linear slopes of change is questionable.

    If the findings reflect initial differences in the mean scores of welders, welder helpers, and non-welding workers, they would suggest that welders are at higher risk of parkinsonian signs. However, it is important to keep in mind that the welders in this study represent a volunteer sample with a participation rate of about 50 percent. If the welders who chose not to participate in the study did so because they had a lower rate of symptomatology than those who did, the initial level of parkinsonian signs in “welders” would be overestimated. The non-welding comparison group selected by the union without input from the investigators could well have been selected on the basis of good health and lack of parkinsonian features, assuring a low mean UPDRS score at baseline and low rate of progression at follow-up. The role of selection bias in producing the differences in baseline performance on the UPDRS3 in this study needs to be considered. 

  2. This study by Racette and colleagues revealing a dose-response relationship between exposure to manganese containing welding fumes and parkinsonism makes important contributions to the fields of neurotoxicology and industrial hygiene. It is, however, important to recognize that while these findings provide additional support for the role of manganese in parkinsonism, these data do not indicate that manganese causes idiopathic PD per se. That said, it is nevertheless scientifically plausible that the changes in neurological function observed in this study could exacerbate those due to idiopathic PD in workers exposed to welding fumes. Stabilization or even improvement in neurological function is expected to occur with cessation of exposure in those subjects with manganese-induced parkinsonism, whereas symptoms would be expected to continue to progress despite cessation of exposure in those patients with idiopathic PD. 

Make a Comment

To make a comment you must login or register.


News Citations

  1. Is Manganese a Risk Factor for PD?

Paper Citations

  1. . Manganese poisoning in Moroccan miners. Br J Ind Med. 1955 Jan;12(1):21-35. PubMed.
  2. . Basal ganglia intensity indices and diffusion weighted imaging in manganese-exposed welders. Occup Environ Med. 2012 Jun;69(6):437-43. Epub 2012 Mar 23 PubMed.
  3. . Increased risk of parkinsonism associated with welding exposure. Neurotoxicology. 2012 Oct;33(5):1356-61. Epub 2012 Sep 3 PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Dose-dependent progression of parkinsonism in manganese-exposed welders. Neurology. 2016 Dec 28; PubMed.
  2. . Understanding of the role of manganese in parkinsonism and Parkinson disease. Neurology. 2016 Dec 28; PubMed.