We do not yet have enough information about the specific variants contributing to AD to use genetic data for clinical risk prediction. These findings conclude that despite interest in genetic testing, and availability of testing through direct-to-consumer avenues, genetic testing for AD is not yet ready to be applied in a clinical setting. This study suggests that expanding the number of replicated variants associated with AD would account for a greater portion of the genetic variance for AD and therefore improve risk prediction. Because AD also has a substantial unique environmental etiology in addition to genetic, a prediction tool based on genetic information alone would not have the highest AUC; the addition of environmental factors would account for more of the variability in AD and therefore a model that takes into consideration both could have better predictive ability. Data simulations in our study show that adding environmental effects could potentially raise the predictive accuracy to 0.95 (Maher et al., in preparation). While genetic information may be of limited clinical validity at the moment, as we continue to identify genes successfully, and incorporate information from both genetic and environmental risk factors, there is potential for future clinical utility.
