Article Text

PDF
Abstracts from the Association of British Neurologists Annual Meeting 2011
120 An algorithm to identify individuals at high-risk of Parkinson's disease in the community
  1. A Noyce,
  2. J Bestwick,
  3. C H Hawkes,
  4. C H Knowles,
  5. J Hardy,
  6. A J Lees,
  7. L Silveira-Moriyama,
  8. G Giovannoni,
  9. A Schrag
  1. Reta Lila Weston Institute of Neurological Studies, Wolfson Institute of Preventive Medicine, Barts and the London School of Medicine and Dentistry, UK
  2. Blizard Institute of Cell and Molecular Science, Barts and the London School of Medicine and Dentist, UK
  3. Institute of Neurology, Royal Free Campus, University College London, UK

Abstract

Studies show that disorders such as hyposmia, constipation and disturbances of sleep and mood precede the development of PD. Individually, the elevated risk of PD in the presence of these variables is modest. No study to date has determined their combined ability to predict PD. We conducted an extensive literature review of published studies of risk factors for PD. A MeSH terms search of the Pubmed database in November 2010, identified 3719 English-language articles. Articles were excluded if they reported on non-motor features in established PD, atypical and genetic forms of parkinsonism or disease management. We excluded reviews and those papers that did not report original data. 148 articles remained. These were reviewed to determine if they contained results from which likelihood ratios could be derived. The references were reviewed for any studies that might not have been captured. Meta-analysis methods were used to combine results of risk factors reported in multiple studies. The algorithm uses Bayes' theorem to estimate risk of developing PD (prior age related risk multiplied by the combined likelihood ratio for the risk factors). Risk factors are assumed to be independent. Performance of the algorithm (sensitivity and specificity) was determined by simulation using the age and sex distribution of the UK population. Our preliminary analysis of the algorithm shows that it could predict the future development of PD with 71% sensitivity and 85% specificity. Combining variables could identify individuals with increased risk of future PD better than individual risk factors. Such an algorithm could be valuable in screening for those at risk of PD in the community.

Statistics from Altmetric.com

Footnotes

  • Email: alastair.noyce{at}mac.com

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.