[Skip to Navigation]
Views 2,984
Citations 0
Original Investigation
October 25, 2021

Long-term Risk of Parkinson Disease Following Influenza and Other Infections

Author Affiliations
  • 1Department of Population Medicine at Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, Massachusetts
  • 2Department of Clinical Epidemiology, Department of Clinical Medicine, Aarhus University Hospital, Aarhus, Denmark
  • 3The Danish Clinical Quality Program, National Clinical Registries, Aarhus, Denmark
  • 4Department of Nuclear Medicine and PET, Aarhus University Hospital, Aarhus, Denmark
  • 5Clinical Excellence Research Center, Stanford University, Stanford, California
  • 6Department of Epidemiology and Population Health, Stanford University, Stanford, California
  • 7Department of Neurology and Neurological Sciences, Stanford University, Stanford, California
JAMA Neurol. Published online October 25, 2021. doi:10.1001/jamaneurol.2021.3895
Key Points

Question  Is influenza or other infection associated with Parkinson disease?

Findings  In this case-control study of all Danish citizens with Parkinson disease between 2000 and 2016, unlike most other infections, influenza was associated with Parkinson disease more than 10 years after infection.

Meaning  Influenza infection may increase the long-term risk of developing Parkinson disease.


Importance  Influenza has been associated with the risk of developing Parkinson disease, but the association is controversial.

Objective  To examine whether prior influenza and other infections are associated with Parkinson disease more than 10 years after infection.

Design, Setting, and Participants  This case-control study used data from 1977 to 2016 from the Danish National Patient Registry. All individuals with Parkinson disease, excluding those with drug-induced parkinsonism, were included and matched to 5 population controls on sex, age, and date of Parkinson diagnosis. Data were analyzed from December 2019 to September 2021.

Exposures  Infections were ascertained between 1977 and 2016 and categorized by time from infection to Parkinson disease diagnosis. To increase specificity of influenza diagnoses, influenza exposure was restricted to months of peak influenza activity.

Main Outcomes and Measures  Parkinson disease diagnoses were identified between January 1, 2000, and December 31, 2016. Crude and adjusted odds ratios (ORs) and 95% CIs were calculated by conditional logistic regression overall and stratified by time between infection and Parkinson disease (5 years or less, more than 5 to 10 years, more than 10 years).

Results  Of 61 626 included individuals, 23 826 (38.7%) were female, and 53 202 (86.3%) were older than 60 years. A total of 10 271 individuals with Parkinson disease and 51 355 controls were identified. Influenza diagnosed at any time during a calendar year was associated with Parkinson disease more than 10 years later (OR, 1.73; 95% CI, 1.11-2.71). When influenza exposure was restricted to months of highest influenza activity, an elevated OR with a wider confidence interval was found (OR, 1.52; 95% CI, 0.80-2.89). There was no evidence of an association with any type of infection more than 10 years prior to Parkinson disease (OR, 1.04; 95% CI, 0.98-1.10). Several specific infections yielded increased odds of Parkinson disease within 5 years of infection, but results were null when exposure occurred more than 10 years prior.

Conclusions and Relevance  In this case-control study, influenza was associated with diagnoses of Parkinson disease more than 10 years after infection. These observational data suggest a link between influenza and Parkinson disease but do not demonstrate causality. While other infections were associated with Parkinson disease diagnoses soon after infection, null associations after more than 10 years suggest these shorter-term associations are not causal.

Add or change institution
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    1 Comment for this article
    Influenza Infection and Parkinson's Disease
    Mahyar Etminan, PharmD, MSc (Epi) | University of British Columbia
    In their study Cocoros1 and colleagues undertook a case-control study using a large Danish population database and found an increase in the risk of Parkinson’s Disease (PD) and influenza with an adjusted odds ratio (OR) of 1.73 (95%CI:1.11-2.71) and 1.91 (95%CI:1.14-3.1) for the 10 and 15 year periods respectively. The study covered a 39 years time span (1977-2016) which might subject the study to a number of time related biases.

    In the study each case was matched to five controls by age, sex and the time of PD diagnosis. Given the long time span of the study, this
    manner of control selection may be subject to time-window bias2. Time window bias is a bias that pertains to case-control studies where the time period for exposure measurement (influenza diagnosis) can be variable between the cases and controls which might lead to biased estimates2. For example, in this study it is possible for a case to be diagnosed in 2017 and have 10 years of data for influenza diagnosis ascertainment. Theoretically this case could have been matched to controls who only had 3 years of data in the database and thus a much shorter time span and in turn opportunity to be diagnosed with influenza. This can lead to an over representation of unexposed controls potentially leading to a higher magnitude of the OR.

    Moreover, the study should have also controlled for calendar time bias as during an approximately 40-year study period influenza infection rates have varied over time as demonstrated in Figure 1 of the study. Similarly, diagnosis, type of treatment and access to treatment for PD might have also changed over this very long period. Thus, calendar time should have been controlled for in the study design to minimize this bias.

    Given the long latency of PD and the time varying nature of influenza, time dependent biases become important variables that need to be properly addressed especially for causal questions such as this one. We suggest that a retrospective cohort study that addresses influenza infections as a time varying exposure and can also control for other competing events might be a more optimal study design to answer this question.

    Mahyar Etminan PharmD MSc
    Department of Ophthalmology and Medicine
    University of British Columbia

    Ali Samii MD
    Department of Neurology
    University of Washington