Do pesticides cause Parkinson's?

Does pesticide exposure cause Parkinson's?

A modest statistical link exists between pesticides and Parkinson's disease, but it weakens under scrutiny. A major systematic review of agricultural risks combined 11 high-quality Tier 1 studies and 38 lower-quality Tier 2 studies. This review calculated a relative risk, which is the ratio of disease probability in an exposed group compared to an unexposed group, of 1.22 using a model assuming a single true effect. When using a model allowing the effect to vary across populations, the relative risk rose to 1.56.

Adjusting for publication bias severely reduces these risk estimates. Publication bias is the tendency of journals to only publish positive results. When researchers adjusted for this bias, the relative risk fell to 1.14 in the single-effect model and became a statistically non-significant 1.11 in the variable-effect model. A separate meta-analysis of 173 studies reported an odds ratio, which measures the strength of association, of 1.77 for general pesticide use and 1.26 for farming.

Real-world exposures are too correlated and misclassified to prove direct causation. There is a 0.66 correlation between herbicide and insecticide use, making it impossible to isolate single compounds. Furthermore, people who work with pesticides are often exposed to well water, which has an independent association of 1.21. Finally, a pioneering Canadian study by researcher André Barbeau found a 20 percent discrepancy in patient counts when comparing record-based data to direct clinical examinations, showing how frequently the disease is misclassified.

Which specific chemicals show the strongest links?

Epidemiological findings for paraquat are deeply contradictory across different study designs. A 2018 meta-analysis of 13 case-control studies with 3,231 patients and 4,901 controls found a 64 percent increased chance of disease, yet its only included prospective cohort study showed no significant link. A 2025 meta-analysis pooling 14 studies with 4,731 cases and 205,129 controls found a significant risk of 1.47 in case-control studies, but a protective effect with an odds ratio of 0.81 in prospective cohort studies, which track healthy people over time. A 2026 preprint, which is an unverified paper that has not yet undergone peer review, showed that six international regulatory bodies reached highly divergent conclusions using the same data.

Rotenone and organochlorines show biological activity but lack robust human evidence. While rotenone is widely used to damage mitochondria and trigger alpha-synuclein clumping in animals, there is a general lack of human epidemiological data specifically isolating it. For organochlorines, a 2024 meta-analysis of 17 case-control studies found a risk increase in Asia with an odds ratio of 1.54 across seven studies measuring blood residues, but no significant association in North America or Europe. Furthermore, the precise biological mechanisms by which organochlorines might damage dopamine neurons remain completely unknown.

Does living near sprayed fields carry the same risk as working with them?

Direct occupational handling of pesticides carries a much stronger statistical link to Parkinson's than living near sprayed fields. In a systematic review of environmental toxins, researchers calculated that occupational exposures carried a high risk with an average odds ratio of 3.82, compared to residential exposures at 2.36. Similarly, a systematic review of rural risk factors found that well-water consumption had a weak relative risk of 1.02 under a model assuming a single true effect, while a systematic review of 23 good-quality studies of direct occupational handling reported a higher risk of 1.67.

Measuring lifetime environmental exposure to drifting chemicals remains highly inaccurate. Most studies do not use direct measurements, relying instead on subjective memory. In a systematic review of metal exposure that included 83 case-control studies and 5 cohort studies, investigators in 69 of those studies had to rely on self-reported histories or biological samples taken after a diagnosis had already occurred. Additionally, a 2026 preprint, which is a paper that has not undergone peer review, notes that the US Environmental Protection Agency reopened its paraquat assessment in 2025 because of new data on how the chemical turns into gas in the air.

Why do most people exposed to pesticides never develop the disease?

The theory that genetic differences explain why only some exposed people develop the disease is unproven. The prevailing idea is that certain individuals inherit variations in detoxification genes, causing their bodies to fail to clear chemicals. However, the evidence supporting these gene-environment interactions is small, retrospective, and frequently contradictory.

Global analyses of detoxification genes regularly contradict small, positive studies. One retrospective case-control study of 254 patients and 169 controls linked missing GSTM1 and GSTT1 genes to pesticide-related Parkinson's. However, a systematic meta-analysis of 22 publications containing 23 studies across multiple ethnic populations showed no significant association for the GSTM1 null genotype, finding an overall combined odds ratio of 1.06. Similarly, a meta-analysis of the MDR1 transporter gene found a weak association for one genetic variant but no significant relationship for another.

Data on other potential genetic filters are thin and incomplete. A meta-analysis of variations in the NAT2 processing gene concluded they do not support an association with Parkinson's risk outside of some Asian populations, noting that most reviewed studies had data gaps. Furthermore, a link between pesticide exposure and the protective PINK1 gene rests on a single retrospective study of just 48 early-onset patients and 61 controls, and no prospective cohort study has ever been run to confirm this finding.

If you already have Parkinson's, does past exposure change how the condition progresses?

We do not know whether a history of pesticide exposure changes how Parkinson's progresses after diagnosis. A systematic review of 11 reviews and 93 primary studies found that no systematic reviews and only four primary articles addressed the factors that influence disease progression, identifying blood urate as the only linked factor. Another systematic review that specifically searched for environmental factors affecting progression found only eight eligible studies, none of which investigated pesticide exposure.

The statistical link between pesticide history and post-diagnosis outcomes is limited to mortality. A retrospective cohort study of 233 Parkinson's patients found that occupational pesticide exposure doubled the mortality rate, but this study did not track the speed of motor decline or the severity of non-motor symptoms like sleep issues. A 2026 preprint, which is an unverified paper that has not undergone peer review, analysed 9,887 participants in Latin America and confirmed that occupational exposure increases the initial risk of developing the disease, but it did not examine how patients fared after their diagnosis.

Nobody has tested whether past pesticide exposure alters a patient's response to standard medications. There are no clinical trials or observational studies evaluating if a history of pesticide exposure makes standard therapies like levodopa less effective. The only available clinical information is a case report of a single 54-year-old man who developed symptoms after an acute glyphosate accident, which cannot tell us how routine, historical exposure affects medication response in typical idiopathic cases.

What this does not show

Major reviews do not prove that individual pesticide exposure causes Parkinson's. A meta-analysis of 13 studies involving 4,731 cases and 205,129 controls found no overall relationship between paraquat and the disease, even showing a lower rate in the occupational subgroup. In a separate systematic review, the only high-quality Tier 1 study, conducted by Firestone and colleagues, found no association between paraquat and the disease, reporting a relative risk of 0.9. Additionally, pooled analyses are highly fragile; in one review, a single study by Brighina and colleagues accounted for over 75 percent of the statistical weight in five out of six models.

Several prominent papers in Parkinson's research have been retracted. A meta-analysis by Wang and colleagues evaluating integrated Chinese and Western medicine was retracted, and another meta-analysis by the same lead author on wound infections in joint arthroplasty was also retracted. In addition, a systematic review of zonisamide by Kong and colleagues and a clinical trial by Lu and colleagues on subthalamic nucleus volume were both retracted. For clinical interventions, a randomised trial on probiotic administration in 60 patients by Tamtaji and colleagues was retracted, alongside a trial on combining pramipexole and levodopa by Huang and colleagues. Finally, a preprint genomic screen of 755 early-onset patients was retracted to protect patient privacy.

No clinical trials have evaluated treatments specifically for pesticide-exposed populations. Furthermore, regulatory bodies remain deeply divided. The Australian Pesticides and Veterinary Medicines Authority kept paraquat registered in June 2026, which was detailed in a preprint that has not undergone peer review.

This review provides the statistical limits and unresolved scientific debates surrounding historical pesticide exposure, but it cannot map these general risks to your personal genetic profile. A clinician can help you interpret how your specific environmental history and symptoms fit into a personalised, long-term care plan.