Drug-wide study associates 31 drug classes with the risk of Parkinson’s disease

Julia Romanowska

University of Bergen

11/15/23

Take-home message

Use of certain drugs can change risk of developing PD even up to 10 years after taking this drug.

How did we come to this conclusion?

• Norwegian Prescription Database
• hypothesis-free screening
  (Cox regression)
• ATC-code level 2
  (C09 = agents acting on renin-angiotensin system)

• NorPD: entire dataset, more than 600 million prescriptions from ca.4.5 million Norwegians, in the period 2004-2019
• Cox regression with time-dependent exposure and age as time scale, adjusted for sex and education (as a proxy for socio-economic status)
• grouping the drugs

10 drug classes PD-risk, 21 drug classes PD-risk

significant associations at FDR<0.05

Can we trust the results?

PROBLEMS:

• long prodromal phase
• long-term vs. short-term drug usage

• PD is a slow-developing disease with various symptoms appearing in a random sequence over 10 or more years before the final diagnosis can be set
• some drugs can be taken sporadically, some need to be taken longer, some people take a smaller dose of a drug, some take larger - how to account for that?

Can we trust the results?

PROBLEMS:

• long prodromal phase
• long-term vs. short-term drug usage

SOLUTIONS:

• time-lag analyses
• dose-response analyses

We have conducted several sensitivity analyses, most importantly:

• time-lag analyses, where the exposure period was shortened for everyone by 3, 5, 8, or 10 years
• dose-response analyses, where the exposure was modeled as a categorical variable with quartiles of the number of prescriptions taken

Drug classes retaining the effect throughout time and doses

Drug classes retaining the effect throughout time and doses

Conclusion

drug classes PD-risk

help identify patients early
e.g., urologicals, psychoanaleptics

drug classes PD-risk

might be repurposed as new treatment
e.g., drugs acting on renin-angitensin system, drugs for obstructive lung disease

Long way to go - all the RCTs have so far failed, but one problem might be that patients involved in these had too advanced stage of PD for the drug to be able to reverse the brain damage.

Check it out yourself!

https://jromanowska.github.io/Norwegian_drug_screening_ATC_level2/

Julia Romanowska, Kjetil Bjornevik, Marianna Cortese, Julia A Tuominen, Magne Solheim, Asieh Abolpour Mofrad, Jannicke Igland, Clemens R Scherzer, Trond Riise,
“Association Between Use of Any of the Drugs Prescribed in Norway and the Subsequent Risk of Parkinson Disease: A Drug-wide Association Study”,
Neurology, 2023, DOI: 10.1212/WNL.0000000000207899