Keynote

Genetic Screening for Susceptibility to Disease

  1. Helen M Wallace

Published Online: 15 SEP 2009

DOI: 10.1002/9780470015902.a0021790

eLS

eLS

How to Cite

Wallace, H. M. 2009. Genetic Screening for Susceptibility to Disease. eLS. .

Author Information

  1. GeneWatch UK, Buxton, UK

Publication History

  1. Published Online: 15 SEP 2009

Abstract

Claims of current and future benefits from genetic susceptibility screening are misleading because genetic variants are poor predictors of most diseases in most people and biology is complex. To date, the single and combined effects of common genetic variants on disease risk do not meet medical screening criteria for the general population.

Nevertheless, numerous commercial companies are attempting to establish a market in such tests, often combined with the personalized marketing of associated products or services. Significant political and financial commitments have been made to implementing genetic susceptibility screening in whole populations. These commitments reflect the historical promotion of genetic ‘prediction and prevention’ of disease by vested interests, and rest on a number of implausible biological assumptions. An independent assessment of the likely costs and benefits of this health strategy is long overdue. Regulation of the clinical validity and utility of genetic susceptibility tests is also needed.

Key concepts

  • Strategies for preventive health are influenced by commercial and political considerations.

  • The single or combined effects of common genetic variants on disease risk do not meet medical screening criteria for the general population.

  • Claims of future improvements in prediction depend on oversimplified models and implausible assumptions.

  • Heritability calculations assume that the effects of interactions, chance, society and choice on the variance of a trait are negligible or nonexistent.

  • There are always limits to the predictability of complex systems, particularly when the underlying physics or biology is poorly understood.

  • It is normal for highly parameterized models to give poor predictions.

  • Data-mining tends to lead to well-calibrated models with little or no predictive value.

  • Successful personalized marketing is not the same as delivering benefits to health.

  • The lack of a premarket assessment of clinical validity and utility means that most genetic susceptibility tests currently on sale are misleading to consumers.

Keywords:

  • polygenic;
  • susceptibility;
  • screening;
  • heritability;
  • prediction