DNA Dilemma Part Two: What Do Genetic Tests Show?

courtesy of Mary Carmichael

To start my examination of whether it's worthwhile to get a direct-to-consumer (DTC) genetic test, I wanted to know exactly what I should and shouldn't expect to learn from the results. How much would the analysis tell me about my health risks compared to the average person's? Were studies linking genes to diseases really translatable into useful predictions? And—on a more light-hearted note—what might ancestry testing tell me? I sought advice from five people (including the founder of the biggest DTC genetic testing company) who I thought would have different takes on some of the questions. They all seemed to agree on at least one thing: while the tests can be educational and interesting, their medical value is debatable at best. Here is what five experts told me I could learn:

Jeffrey Barrett

Statistical geneticist, Wellcome Trust Sanger Institute
"Common conditions [some researchers hope] to predict (such as diabetes or schizophrenia) aren't caused by a single catastrophic gene defect, but are instead subtly influenced by a combination of many different genes and environmental factors like diet and exercise. Progress in understanding the genetic part of that equation has accelerated rapidly in the last four years.

Unfortunately, the gene variants discovered so far barely nudge someone's overall risk, typically increasing it by a factor of 1.1 to 1.5. The studies typically report 'relative risk,' or the increased chance someone with a particular genetic variant has of getting sick compared to the background rate of that disease in his community. Translating this information to a meaningful personal prediction can be tricky, because the background rate can vary widely around the world.

Nevertheless, the incorporation of additional medical information and new discoveries aided by ever-cheaper genome sequencing could yet produce useful individual risk predictions."

Read a more in-depth version of Barrett's answer at Genomes Unzipped.

Mary's take: Barrett's perspective was a great reality check. Often, the media excitedly reports findings as "scientists discover genes for [insert common condition here]," and it's easy to make the leap from that to "if I have those genes, I'll get that condition." But the genes can have wimpy effects, and the typical health risks in the populations studied may be different than those facing a given consumer. What Barrett seemed to be telling me was that I shouldn't stop at the test results: I should also look critically at the studies they were based on. This sounded appealing, if very time-consuming.

Anne Wojcicki
President and Co-founder, 23andMe

"Genetics can impact whether you'll need more or less of a medication compared to most people, or whether you might be at increased risk for serious side effects. But DTC genetic testing can't replace your doctor when it comes to making decisions about how best to treat you. There are many factors other than genetics that must be taken into consideration.

You can also learn about mutations that could impact your children. If two people are carriers of certain disease-causing mutations, each child they have together will have a 25 percent chance of being born with the disease. 23andMe tests for mutations associated with cystic fibrosis, Tay-Sachs disease, sickle cell anemia, and many other conditions. Knowing a child has a chance of being born with one of these diseases can facilitate early intervention and improve outcomes. But remember: All screening tests necessarily cover only a subset of the possible mutations for each disease. You could still carry a mutation even if your results from a DTC genetic testing company come back negative.

Common DNA variations found within all of us have been linked to increased risk for things like diabetes, cancer, and autoimmune conditions. However, the genetic variations identified to date explain only a small part of a person's risk for each disease. Unknown genetic factors, as well as environment, are also important. Additionally, people from ethnic minorities get limited information from DTC genetic testing. Through its research, 23andMe is working to change this.

There's also fun stuff: You can trace your ancient ancestry, find distant cousins, and learn the science behind physical traits like eye color and earwax type."

Read a longer version on the 23andMe blog.

Mary's take: I could see how the drug-response results could be medically useful—some doctors already consider it standard of care to test for patients' responses to the drugs Plavix and Coumadin, which are anti-clotting medications. On the other hand, I'm not actually taking those drugs. Carrier results, even if they were incomplete, would be interesting to learn (although they wouldn't tell me much about my children's futures unless I also convinced my husband to get tested). As for common disease risks, I was impressed with 23andMe's willingness to be circumspect. But I wondered if other consumers were getting the same message. The company's Web site exhorted customers to "live well at any age by understanding your health risks." Hadn't Barrett just told me that scientists don't understand my health risks? Finally, there was the "fun stuff," which was a big draw for me. Maybe taking the test would be a little like having a kid's chemistry set: good geeky entertainment, as long as I didn't use it to blow anything up.

Muin Khoury
Director of Public Health Genomics
for the CDC

"You as a consumer [of a DTC test] will get a printout that says, 'You have the following sets of gene variants, and your risk of this disease is X percent or Y percent.' Those numbers are not trustworthy at this point. They vary from one company to another, as we saw in the GAO report. And the science is changing. If you're trying to come up with your risk profile for type II diabetes, for instance, someone may tell you, 'On the basis of these five genes I measured, you have increased risk.' But if you add two more genes, the risk could go down. Whereas if you know somebody's age, gender, race, BMI, and family history, you can already predict with almost 90 percent certainty if they will get type II diabetes.

Let's assume the report tells you you're at slightly increased risk from the average. Are you supposed to exercise to decrease your risk? You should do that anyway. If knowledge of the genome today doesn't tell me anything more than I already know, why would I spend several hundred bucks?

I'm not telling people not to buy the tests – I'm just telling them the information may not be all they think it is."

Mary's take: Khoury recently made an official statement—the CDC's first—on DTC genetic tests, saying they aren't "ready for prime time." The concerns he laid out for me left me appropriately skeptical, but they weren't enough to convince me not to take the test. Genetics is very much a work in progress and that risk assessments will change as the science does—that's something everyone agrees on. But lots of medical measurements are imprecise indicators of health, including BMI (not to mention race). Ultimately, Khoury's answer to my question left me with another one: How much do scientists have to know before these tests are ready for prime time?

Steven Murphy
physician specializing in personalized medicine and vocal DTC critic

"I have seen probably more patients with these types of tests than just about any clinician out there, so I can tell you what the patients ask and what I tell them. (1) You will not learn what you will die from. In fact, most tests won't tell you, whether ordered via Amazon or through your doctor. (2) You will not learn what diet is right for you. We do know some people may benefit from some types of diets to lower cholesterol and lose weight, but the science to accurately predict genes and diet is not ready to bring to market. (3) You will not learn if you are of the lost tribes of Israel. The J Haplogroup analysis you get is only a small window into heritage.

The results for positive carrier tests are just as valid as those ordered by a doctor in some instances. But they may be better at ruling in than ruling out. If it says you are not a carrier, don't trust the results. If it says you are, you can double check with a doctor. You may learn you carry a rare mutation putting you at significantly increased risk of breast or ovarian cancer. Are you ready for what to do with those results if you had them?" Read a longer version at Murphy's blog.

Mary's take: Murphy may have seen patients asking "what they'll die from" based on a DTC test, but I myself wouldn't ask such a silly question. I appreciated his concern that I might not be ready to find out about BRCA, the breast and ovarian cancer mutation, and so, apparently, did the test makers. Some potentially life-altering results are "locked" in many DTC reports. You're supposed to specifically decide you're ready to see them before you actually do. Mirabile dictu, I started to believe I was reaching a consensus among my contributors: you won't get much good out of the test if you don't actively think about the results.

Blaine Bettinger
lawyer by day, biochemistry Ph.D. and genetic genealogy expert by night

"Everyone has both a genetic tree and a genealogical tree
. Your genealogical tree includes all your ancestors. Your genetic tree is comprised of a small, random assortment of ancestors who contributed DNA to your genome. Ancestry testing applies only to these people. With this limitation in mind, let's look at what it offers.

An especially exciting product is the admixture analysis, which determines for each segment of DNA whether it was likely to have been inherited from ancestors in Africa, Asia (including Native Americans), or Europe: my 23andMe test suggests that I am 97.89 percent European, 1.84 percent Asian, and 0.27 percent African. Results can differ from company to company based on which genetic variants, reference populations, and algorithms are used. My deCODEme analysis suggests I am 87 percent European, 9 percent Asian, and 4 percent African. You can also identify and connect with "genetic cousins," i.e., people who inherited DNA from the same common ancestor."

Read more on Bettinger's blog.

Mary's take: I'm my family's official record-keeper and I can track my ancestors back to 1638, when one of them first showed up in America. In other words, I find traditional genealogy entertaining—so wouldn't it be cool to augment what I already know? Yet even with ancestry testing there are ethical and medical questions to consider.

I tucked all my essays and interview notes away and decided to tackle the topic from a new angle, which I'll explore tomorrow: whether or not I could trust the results of a DTC test, and whether I should worry about what I might find out.

Follow Mary on Twitter as she makes her decision, or add your opinion using the #DNADilemma hashtag.

DNA Dilemma: The Complete Series

Monday: Is Now the Time to Test?
Tuesday: What Can I Learn From This Test?
Wednesday: How Meaningful Are The Results?
Thursday: Should These Tests Be On the Market?
Friday: My Genome, Myself: I Make My Decision.
Recommended Reading: Further Articles, Essays, and Websites
DNA Dilemma FAQs: You're Doing What, Exactly?