Parkinsonism was linked with chronic traumatic encephalopathy (CTE) in athletes, an analysis of cross-sectional data suggested.

Nearly one in four (24.7%) of 481 deceased contact sports participants with CTE had parkinsonism symptoms before they died, according to Ann McKee, MD, of Boston University, and co-authors.

Most deceased athletes with parkinsonism and CTE (75.9%) had an unusual pathology related to CTE and not the typical Lewy body pathology often seen in Parkinson's disease, McKee and colleagues reported in JAMA Neurology.

"This study establishes a new link between playing contact sports, CTE, and the development of parkinsonism," co-author Thor Stein, MD, PhD, also of Boston University, told MedPage Today. "It highlights that CTE pathology, not Lewy body pathology, is the primary driver of parkinsonism symptoms in most cases."

Compared with other athletes with CTE, those with parkinsonism had a more severe CTE stage and more nigral pathology. In the substantia nigra, CTE participants with parkinsonism symptoms were more likely to have:

• Neurofibrillary tangles (42.7% vs 29.9%, P=0.01)
• Neuronal loss (52.1% vs 17.1%, P<0.001)
• Lewy bodies (24.1% vs 5.8%, P<0.001)

CTE is defined at autopsy by hyperphosphorylated tau protein deposits within neurofibrillary tangles distributed around blood vessels and at the depths of the cortical sulci; it's associated with repetitive head impact exposure. Parkinsonism is a clinical motor dysfunction syndrome characterized by bradykinesia, rigidity, and resting tremor.

For over 100 years, parkinsonism has been a recognized symptom in athletes like boxers who often are decades removed from repeated head hits.

"Historical case reports did not have the benefit of clinicopathological correlation of parkinsonism in individuals with repetitive head impact," observed Breton Asken, PhD, of the University of Florida in Gainesville, and co-authors in an accompanying editorial.

"We now better understand the highly specific association of repetitive head impact with CTE, a neurodegenerative tauopathy, but there is a growing appreciation for the spectrum of neuropathological consequences linked to repetitive head impact beyond or combined with CTE," Asken and colleagues wrote.

McKee and co-authors studied autopsy data from male brain donors with CTE and no other significant neurodegenerative disease from the UNITE brain bank between July 2015 and May 2022. Postmortem informant interviews, online surveys, and medical records also were evaluated, including specific information about bradykinesia, resting tremor, rigidity, and shuffling gait.

In this sample, American football was the more frequent sport that participants with parkinsonism played (90.8%). Men with parkinsonism were older when they died than those without parkinsonism (mean age 71.5 vs 54.1 years).

Larger proportions of participants with parkinsonism had symptoms of dementia (87.4% vs 29.0%), probable rapid eye movement sleep behavior disorder (43.7% vs 16.0%), and visual hallucinations (37.8% vs 14.1%) than those without parkinsonism (P<0.001 for all).

Years of contact sports participation -- a proxy for repetitive head impacts -- were associated with nigral neurofibrillary tangles (adjusted odds ratio [AOR] 1.04, 95% CI 1.00-1.07, P=0.03) and neuronal loss (AOR 1.05, 95% CI 1.01-1.08, P=0.02), McKee and colleagues said. Nigral neuronal loss (AOR 2.61, 95% CI 1.52-4.47, P<0.001) and Lewy bodies (AOR 2.29, 95% CI 1.15-4.57, P=0.02), in turn, were associated with parkinsonism.

Substantia nigra neuronal loss was associated with nigral Lewy bodies (AOR 4.48, 95% CI 2.25-8.92, P<0.001), nigral neurofibrillary tangles (AOR 2.51, 95% CI 1.52-4.15, P<0.001), and arteriolosclerosis (AOR 2.27, 95% CI 1.33-3.85, P=0.002). Overall, nigral neurofibrillary tangles and neuronal loss mediated the association between years of American football play and parkinsonism in individuals with CTE.

The study had several limitations, the researchers acknowledged. The sample was selective and findings may not apply to other populations. Clinical data were generated mainly by informant-based retrospective review and may have been influenced by recall bias. The study also could not discern drug-induced parkinsonism.