Thank you for that question.
There are two categories of genetic tests with respect to this question. One of them is the single gene test, such as the test for Huntington's disease, where there is a definitive test result: you either will or will not develop that disease, depending on that test result. Those are very reliable tests that we use in the context of our medical relationship with patients.
The second category, which you alluded to, are genomic variants, the little variations in each of our 20,000 or so genes that may contribute a small amount toward what we call a “complex disease”, such as heart disease. There are probably many genomic variants that contribute to the predisposition to heart disease, and that aspect of genetic testing is still very much emerging.
The best example I can give you of the unreliability of that information is that direct-to-consumer genetic test companies are using those variants to offer patients. Patients swab their mouth, send it off in a kit to a company, and get back a report about all the variants in their genome and ostensibly what their risk is of having heart disease, Parkinson's, and various other things.
If you take the same person's swab and send that to six different direct-to-consumer test companies, you will get six completely divergent results. You will get one company saying that the person is at high risk for heart disease, another saying they're at average risk, and another saying they're at low risk. This is published. There are published studies, as well this being anecdotal.
In fact, I just read a published study last night that speaks to this very issue. It is extremely dangerous for insurance companies to base their assessments of risk on this kind of data. It is not ready for prime time.