European Human Genetics Conference 2011
Amsterdam, The Netherlands
May 28 - 31, 2011
C14.5 - Risk predictions from direct-to-consumer personal genome testing: What do consumers really learn about common disease risk?
risk prediction; genetic testing; direct-to-consumer genomics
R. R. J. Kalf
, R. Mihaescu
, P. de Knijff
, R. C. Green
, A. C. J. W. Janssens
Erasmus University Medical Centre, Rotterdam, Netherlands,
Leiden University Medical Centre, Leiden, Netherlands,
Harvard Medical School, Boston, MA, United States.
Direct-to-consumer (DTC) companies predict risks of common complex diseases on the basis of genetic markers. Given the low number of markers involved and their small effect sizes, it is unclear whether high-risk groups can be identified. We investigated the risk distributions generated by two DTC companies for 8 diseases.
We simulated genotype data for 100,000 individuals based on published genotype frequencies. Predicted risks were obtained using the formulas and risk data provided by the companies.
The table presents observed and trimmed ranges of predicted risks. The two companies used different formulas to calculate risks. One company predicted risks higher than 100% for 5 out of 8 diseases, which for AMD concerned 1 in 200 individuals. Observed ranges were smaller for the second company, except for Type 1 Diabetes. Predicted risks higher than 50% were frequently observed for company 1, but were exceptions for company 2. When predicted risks of company 1 were calculated using the formulas of company 2, observed ranges were substantially smaller.
Whether companies can currently predict high risks of disease predominantly depends on the method used for their calculation.
Type 1 Diabetes
Type 2 Diabetes
Values are percentages. AMD=Age-related Macular Degeneration; Trimmed range=0.5% outliers removed (two-sided)
The European Society of Human Genetics 2011
The Online Abstract Submission and Invitation System
© 1996 - 2015 Coe-Truman Technologies, Inc. All rights reserved.