Abstract
BACKGROUND: UK Biobank (UKBB) provides extensive genetic, imaging, and health data for ~500,000 participants, enabling studies of prodromal phases of diseases like Parkinson’s disease (PD). However, during analysis, we became concerned about the accuracy of diagnosis timing.
OBJECTIVE: To evaluate the accuracy of PD diagnosis timing in UKBB.
METHODS: We examined PD diagnosis timing using hospital, primary care, death records, and self-reported data. We assessed discrepancies between sources and identified co-occurring diagnoses recorded on the same date as PD.
RESULTS: Among 3979 PD cases, 97% of the 786 participants with both self-reported and electronic health records (EHRs) reported their diagnosis earlier than recorded in the EHR, with a typical delay of 5 to 7 years. Multiple codiagnoses were often logged on the same date, suggesting retrospective or batch data entry.
CONCLUSIONS: Substantial delays in PD documentation may misclassify already diagnosed individuals as prodromal. This introduces significant bias into studies of early disease markers and distorts the timing between risk factors and clinical onset. © 2025 The Author(s). Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.