Gene tests don’t have the answers

Since our parents died, my brothers and I have invented our own ways to spend special occasions. Sometimes we take trips at Christmas, as this seems like a better gift than anything else we could buy. Last Christmas, we went to California’s Big Sur. We ate lavishly, hiked lazily and had a small gift exchange. My older brother’s present, however, surprised us: genetic testing kits.

These direct-to-consumer kits have drawn withering criticism since 2006, when U.S. Government Accountability Office investigators found that the test firms often make medically unproven predictions and provide results so ambiguous that they are meaningless.

All the same, the home test kits, by promising a crystal ball into a person’s medical future, have tantalized many consumers. The Food and Drug Administration estimates that up to 200,000 have been sold in this country.

Given my family’s sad medical history, I felt some of the allure. My father was a kind man but also an alcoholic who drank heavily every night, although he somehow survived to age 83 before dying of liver cancer. My mother succumbed to leukemia at 67, a year after getting her grim diagnosis.


Those memories floated in the back of my mind as I looked over my $199 kit that promised to help me “discover risk factors for 97 diseases,” though I was plenty skeptical.

So what if it showed a genetic predisposition to a disease? Even if you assume that good science was used to produce the test results — a dicey proposition — that’s no guarantee of anything. Scientists still don’t know enough about the complex interplay between genetic variations and an individual’s existing health conditions, environmental exposures and factors such as diet and fitness to make accurate medical predictions.

Still, I obediently followed the directions — spitting into a small, plastic tube, sealing it in an envelope and dropping it in a mailbox. I would have to wait six to eight weeks for the results. In the meantime, I began to do some research.

I learned that the Government Accountability Office had conducted what essentially were sting operations against the direct-to-consumer genetic test industry, first in 2006 and again last year. The industry came out of it looking ugly each time.

In the 2010 investigation, GAO staffers sent the same DNA samples to four companies and received wildly varying results. In one case, a DNA donor was declared by one company to be at below-average risk of getting prostate cancer and hypertension, while other test kit companies found the donor to be at average or above-average risk.

The GAO investigators also discovered that some of the kit companies were making bogus pitches to try to peddle dietary supplements, including claiming that they could use a consumer’s DNA to create personalized supplements to cure diseases.

Aside from the slippery evaluations and shoddy marketing practices, there also is the question of whether consumers — even those lucky enough to buy kits from aboveboard companies — are up to the challenge of interpreting the results. Last month, an FDA official said the agency might require some of the evaluations to be ordered or interpreted by a doctor.

While the agency decides how firmly it will regulate the genetic test companies, people like my brothers and me still are sizing up the results ourselves.


When my results were ready, I was notified via email by 23andMe, the firm my brother purchased our kits from and one of the test companies in talks with the FDA about how to comply with the agency’s requirements. All in all, the test results gave me neither the sense of false assurance nor the unnecessary anxiety that federal authorities warn about.

The results revealed I was at higher-than-average risk for Type 2 diabetes and age-related macular degeneration. (In other words, I have a better-than-average shot of going blind in my old age.)

The test also explored a growing branch of DNA testing — pharmacogenomics — assessing a patient’s likely response to certain drugs. According to the test results, I have an increased sensitivity to warfarin, a blood thinner, but no strong susceptibility to becoming addicted to heroin. Good to know.

I was rated as low risk for melanoma and breast cancer. That would be nice news but, given the history of those conditions in my family, I couldn’t take those readings seriously.


My main frustration with the test was that the results were too wishy-washy.

23andMe, which counts Google and Genentech among its investors, uses a star rating system. Four stars indicates that there are at least two independent studies linking a person’s genetic variation to a disease. But two or three stars means that, essentially, researchers are on the fence. In my case that fuzzy category accounted for 69 out of the kit’s 97 predictions and 11 out of the 19 drug response forecasts.

Perhaps what’s needed, rather than putting these tests into the hands of doctors and geneticists, is simple transparency about what the evaluations can and cannot do. Shortcomings — such as the fact that too little genetic information has been collected on African Americans, Asians and other minorities to put their results in context — should be plainly stated on the package.

I discussed some of these issues with my brothers. My older brother explained he was interested primarily in the test’s ability to provide clues about our ancestry. He also said he was excited about purchasing science at an affordable price. Our kits, at $199, were cheap compared to the $999 that 23andMe charged just a little more than two years ago.


All of this made sense to me.

Still, if the industry’s quest is to provide definitive answers to burning medical questions, it has a long way to go. Perhaps we should have skipped the tests and spent an extra night in Big Sur.

Lilly Fowler is an assistant editor at FairWarning (, a nonprofit, online investigative news organization focused on safety and health issues.