Machine learning nominates drug exposures inversely associated with Parkinson risk
Source check: Source checked. Not independent scientific confirmation.
Finding: LASSO-based screening evaluated historical prescriptions in 40,760 Parkinson cases and 176,395 controls in French health records. Sulfonamide diuretics, especially furosemide, produced the most robust inverse-association signal. Exposure windows preceded diagnosis by years to reduce reverse causation, but confounding and prodromal changes remain possible. These observational hypotheses do not establish prevention, treatment efficacy or a reason to alter medication.
Technology: LASSO logistic regression; repeated sample splitting
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Cross-validated LASSO selects drug/covariate associations in one half-sample; logistic effects are estimated in the other and aggregated across repeated splits.
- Test and result
- Retrospective nested case-control screening with 500 repeated split-selection/estimation cycles and sensitivity analyses.
- Limitation / still unknown
- Causality, residual confounding, reverse causation and preventive/treatment efficacy remain unestablished.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC10087353/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 21 Nov 2026. Review scope: Primary full paper relevant model Methods, study Results and limitations; exact primary front-matter identity and Europe PMC metadata checked..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.