Skip navigation

NSAIDs and risk of Parkinson's disease

 

Clinical bottom line

We can not be sure that use of NSAIDs protects against development of Parkinson's disease. Even if they did, the evidence here suggests that several tens of thousands of people would need to use NSAIDs regularly for one to case of Parkinson's disease prevented.


Reference


H Chen et al. Nonsteroidal anti-inflammatory drugs and the risk of Parkinson disease. Archives of Neurology 2003 60: 1059-1064.


Study

This study was part of the Health Professionals Follow-up Study begun in 1986. It included 51,500 male health professionals aged 40 to 75 years. The Nurses Health Study was begun in 1976 with 122,000 registered nurses aged 30 to 55 years. In both cohorts questionnaires were mailed every two years for update information on potential risk factors for chronic disease, and for occurrence of specific medical events.

Parkinson's disease was ascertained by specialist questionnaire completed by attending neurologist, or by sending a copy of the medical records. A case was definite if the diagnosis was considered definite or probable by a neurologist, or the presence of at least two of three cardinal signs (rest tremor, rigidity, or bradykinesia) in the absence of features suggesting other diagnoses.

NSAID use was determined from specific questions in the participant questionnaires.

Results

Regular (two or more times a week at the beginning of the study) use of non-aspirin NSAIDs was reported by 6.1% of men and 3.7% of women. Users and non-users were the same in terms of smoking status, and alcohol and caffeine consumption.

There were 415 incident cases of Parkinson's disease in over 500,000 person years of observation in men, and 1.7 million person years of observation in women. Thirteen of the incident cases occurred in regular users of NSAIDs (14 per 100,000 person years) and 401 in people who were not regular users (19 per 100,000 person years). This was significantly lower with regular use of NSAIDS after adjustment for age, smoking status, or for a variety of confounders, but only at the 5% level, but 20,000 people would have to take NSAIDs for one case to be prevented. In men, where there was data, only use of NSAIDs for five years or more was associated with a lower incidence of Parkinson's disease, but statistical significance was not achieved, even with adjustment for confounders.

Use of neither aspirin nor paracetamol was not associated with reduced risk of Parkinson's disease.

Comment

Here we have huge cohorts of men and women in impeccably done studies. The result seems to be that regular use of NSAIDs (but not aspirin or paracetamol), taking NSAIDS twice a week for as long as five years, reduces the incidence of Parkinson's disease. The authors give a fine discussion making the case why NSAIDs should be neuroprotective.

We should take some confidence that the incidence of Parkinson's disease found here (19 per 100,000 per year) is the same as that found from a systematic review of incidence studies.

The worry is that only 13 incident cases of Parkinson's disease were found in people taking NSAIDs. Statistical significance was achieved at a bare minimum (relative risk confidence interval of 0.32 to 0.94) and at the 5% level. The result is open to random chance effects. Adding perhaps one or two more cases to the NSAID users would probably make this result not statistically significant. We can not be certain that use of NSAIDs protects against Parkinson's disease. Even if they did, the evidence here suggests that several tens of thousands of people would need to use NSAIDs regularly for one to case of Parkinson's disease prevented.