PREDICT-PD: An online approach to prospectively identify risk indicators of Parkinson's disease

Alastair J. Noyce, Lea R'Bibo, Luisa Peress, Jonathan P. Bestwick, Kerala L. Adams-Carr, Niccolo E. Mencacci, Christopher H. Hawkes, Joseph M. Masters, Nicholas Wood, John Hardy, Gavin Giovannoni, Andrew J. Lees, Anette Schrag*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

52 Scopus citations


Background: A number of early features can precede the diagnosis of Parkinson's disease (PD). Objective: To test an online, evidence-based algorithm to identify risk indicators of PD in the UK population. Methods: Participants aged 60 to 80 years without PD completed an online survey and keyboard-tapping task annually over 3 years, and underwent smell tests and genotyping for glucocerebrosidase (GBA) and leucine-rich repeat kinase 2 (LRRK2) mutations. Risk scores were calculated based on the results of a systematic review of risk factors and early features of PD, and individuals were grouped into higher (above 15th centile), medium, and lower risk groups (below 85th centile). Previously defined indicators of increased risk of PD (“intermediate markers”), including smell loss, rapid eye movement–sleep behavior disorder, and finger-tapping speed, and incident PD were used as outcomes. The correlation of risk scores with intermediate markers and movement of individuals between risk groups was assessed each year and prospectively. Exploratory Cox regression analyses with incident PD as the dependent variable were performed. Results: A total of 1323 participants were recruited at baseline and >79% completed assessments each year. Annual risk scores were correlated with intermediate markers of PD each year and baseline scores were correlated with intermediate markers during follow-up (all P values < 0.001). Incident PD diagnoses during follow-up were significantly associated with baseline risk score (hazard ratio = 4.39, P =.045). GBA variants or G2019S LRRK2 mutations were found in 47 participants, and the predictive power for incident PD was improved by the addition of genetic variants to risk scores. Conclusions: The online PREDICT-PD algorithm is a unique and simple method to identify indicators of PD risk.

Original languageEnglish (US)
Pages (from-to)219-226
Number of pages8
JournalMovement Disorders
Issue number2
StatePublished - Feb 1 2017


  • Parkinson's disease
  • cohort
  • epidemiology
  • prodrome
  • risk factors

ASJC Scopus subject areas

  • Neurology
  • Clinical Neurology


Dive into the research topics of 'PREDICT-PD: An online approach to prospectively identify risk indicators of Parkinson's disease'. Together they form a unique fingerprint.

Cite this