Living close to a dry cleaner is linked a 14% increased risk for Parkinson’s disease (PD) with the strongest association seen in the gait-instability subtype and in specific regional hotspots.

In a nationwide study, the largest of its kind to date, investigators found a dose-dependent association between dry cleaner proximity and PD risk — with the greatest risk found for individuals who lived within 500 ft of the establishments compared with those who lived 4-5 miles away.

In addition, gait disturbance was the clinical feature most strongly linked to dry cleaner proximity, and the regions showing the greatest increased risk included New York City, Southern California, Michigan, and parts of the South.

“Hopefully the study spreads some awareness of what kind of risk this exposure might cause,” study investigator Brittany Krzyzanowski, PhD, now at Atria Research and Global Health Institute, New York City, told Medscape Medical News. 

Krzyzanowski, who was at Barrow Neurological Institute, Phoenix, while the study was conducted, added that there were several unknown factors in the study, including length of stay at the residence.

“These kinds of proximity studies provide early evidence that encourages deeper work in the future,” she said.

The findings were published online on August 5 in npj Parkinson’s Disease.

‘Legacy Contamination’

Previous studies have shown a link between PD and various environmental factors, including pesticides, heavy metals, and air pollution.

A study published online in JAMA Network Open in 2024 by Krzyzanowski and colleagues showed a significant association between increased risk for PD and exposure to particulate matter with a diameter of 2.5 µm or less (PM_2.5) or to nitrogen dioxide. In another study, they found an association between PD risk and ambient trichloroethylene (TCE).

The chlorinated solvent TCE has been used in various industries for such things as cleaning and degreasing and “is recognized as being among the most prevalent groundwater contaminants in the United States,” the investigators noted.

Starting in the 1920s, TCE was used as a dry-cleaning solvent. Although recognition of its toxicity led to replacement by the chlorinated solvent perchloroethylene (PCE) in the 1950s, PCE can degrade and form TCE.

The investigators noted that TCE contamination often occurs alongside PCE beneath dry cleaners and other industrial sites, and that these chemicals can persist in groundwater for decades, forming vapor plumes that seep into nearby homes and businesses. This legacy contamination lingers because PCE and TCE break down slowly and incompletely underground.

Given the density of dry cleaners nationwide and the persistence of TCE and PCE contamination in soil and groundwater, the investigators wanted to examine whether residential proximity to dry cleaners was associated with incident PD risk, and whether that association varied by clinical PD subtype.

The nationwide, population-based, case-control study included 180,982 Medicare beneficiaries at least 67 years of age in 2016-2018 and with PD (mean age, 78.9 years; 54.6% men) and more than 18.8 million individuals without PD who lived within 5 miles of a dry cleaners (non-case controls; mean age, 75.7 years; 56.7% men).

Four PD symptom subgroups were also assessed: tremor, gait disturbance, dementia, and falls. Patients were classified into the tremor subgroup if they carried a code for tremor or abnormal involuntary movement; the gait disturbance subgroup if they carried a code for abnormal gait or mobility; the dementia subgroup if they carried a code for dementia or cognitive decline; and the falls subgroup if they carried a code for an accidental fall.

These subgroups were not mutually exclusive, so patients with multiple co-occurring symptom codes could be counted in more than one subgroup.

Location data on nearly 33,000 US dry cleaners were collected for 2013. Distance to the nearest dry cleaner was calculated using the latitude and longitude of each participant’s zip+4 residence — recorded 2 years before PD diagnosis for cases, or before the reference date for noncases.

Investigators also assessed average annual mean PM_2.5 level from 1998 to 2000, rural-urban commuting area (RUCA), and median household income. 

Dose-Dependent Link

Results showed a dose-dependent link between residential distance to dry cleaners and PD risk throughout the US after adjusting for such factors as age, sex, race, household income, smoking status, RUCA, and average annual PM_2.5.

Participants who lived within 500 feet of a dry cleaner had the highest risk for PD (odds ratio [OR], 1.14; P < .0001) compared with those who lived 4-5 miles away; but even those who lived 2-3 miles away vs 4-5 miles away were at significantly increased risk (P < .01).

The strongest clinical feature of PD associated with dry cleaners proximity was gait instability, with an OR of 1.23 for a residential distance within 500 feet vs 4-5 miles (P < .0001).

Krzyzanowski noted that this finding was not surprising because previous research has shown a link between gait/postural instability and other neurotoxicants, including manganese. It was also found in her earlier TCE and PM_2.5 studies. 

Associations between PD risk and those living within 0.1 miles of a dry cleaners compared with those living more than 0.1 up to 5 miles away were also “strong and significant” in several county cluster areas, including Southern Illinois (OR, 2.50), the Tennessee-Kentucky border (OR, 2.10), Central Pennsylvania (OR, 2.03), north-central Indiana (OR, 1.88), the Dallas-Fort Worth area (OR, 1.50), the greater Detroit region (OR, 1.40), parts of Southern California (OR, 1.16), and New York City (extending into much of New Jersey; OR, 1.13).

These geographic patterns may reflect “differences in vapor intrusion risk related to aging infrastructure, geology, or population density,” the investigators wrote.

Overall, the study “adds to the pile of existing evidence on the association between TCE and Parkinson’s,” Krzyzanowski said.

The investigators pointed to several limitations. Proximity to a dry cleaner does not capture individual-level variability in solvent concentration or building characteristics, and use of location data from a single time point may have introduced exposure misclassification.

They also noted that some establishments identified as dry cleaners may have functioned only as drop-off points, with actual solvent use occurring at a separate facility. However, they note that this type of misclassification would probably weaken rather than strengthen the observed associations — meaning the true effect could be larger than what was reported.

Biologically Plausible

Samuel Goldman, MD, professor in the Division of Occupational, Environmental, and Climate Medicine at the University of California San Francisco, told Medscape Medical News that the study’s use of such a large data sample was a “huge strength.”

Goldman, who was not involved in the study, has conducted previous research on TCE exposure and PD risk.

“This is a relationship that people have been interested in for quite some time and it’s surprisingly difficult to study,” he said. “Based on my research and that of others, I think it’s biologically highly plausible that TCE and PCE releases from any industrial setting can put people at risk who are occupationally or environmentally exposed.”

Goldman said he wasn’t surprised by the dose-response relationship between dry cleaner proximity and PD risk, but he was struck that the association held up consistently across a range of sensitivity analyses.

He was less convinced by the gait-instability finding, however, noting that because the study relied on claims data, it’s difficult to know how accurately the Parkinsonism subtypes were coded.

“I think the strength of that observation wasn’t as compelling as the strength of their overall observation,” he said. “However, increasingly there are data suggesting that people with TCE-associated Parkinson’s disease may have a more aggressive course.”

The investigators and Goldman reported having no relevant financial relationships.