In this issue of JAMA Neurology, Peter and colleagues1 provide a glimpse of the promise and beauty of big data for Parkinson disease (PD) and for neurology in general. By interrogating claims data from more than 170 million people, the authors identified a PD risk increase of approximately 28% among patients with inflammatory bowel disease (IBD), replicating the results of a previous Taiwanese study.2 They then used this link to inspire a virtual repurposing prevention trial. If IBD and PD share a disease-driver network centered around inflammation, could it be that medicines that treat IBD also reduce risk of PD? To conduct a prevention trial for PD from scratch would be incredibly difficult and expensive. It has never been done, although several prevention trials are on the way for Alzheimer disease. Only 11 to 33 annual incident PD cases (“converters”) are expected per 10 000 individuals aged 50 years or older.3 The Parkinson Associated Risk Study,4 for example, screened 9398 eligible participants with a smell test, identifying 669 persons with hyposmia and 23 individuals at high risk (having both hyposmia and a dopamine transporter neuroimaging deficit), 14 of whom converted to PD within 4 years.