For over a decade, it’s been easy to buy a genetic testing kit and mail in a saliva sample for analysis. You might also mail in a sample of urine.

When the answer comes back, you’re likely to hear that you are at risk for an illness, perhaps several. If accurate, this might be useful if it motivates you to take better care of your health.

But despite the hype, most of the time your results cannot definitively tell you whether you will ever have symptoms of any disorder, how severe they might be, and whether the symptoms will progress.

If you’re worried about a particular health risk, and want a gene test, it’s a better idea to get it through a physician, who can interpret the results and help you decide how to advise other members of your family (Horton et al., 2019).  

Here’s our primer on those controversial at-home genetic tests.

What are Direct-To-Consumer (DTC) genetic tests?  

Man collecting genetic test saliva sample
The process at home is simple – swab the mouth, place swab in container, drop in the mail. But is it really worthwhile?

Your genome is made up of thousands of genes that carry information about your body. DTC tests promise to give you useful information based on an analysis of your genes, looking for what are called “variants.”

They usually don’t sequence an entire genome. Instead, they look for specific variants using a method called “SNP-chip genotyping” (Horton et al., 2019). “SNPs,” (pronounced “snips”) stands for single nucleotide polymorphisms. They are the most common kind of genetic variation, and most of them have no known effect – in other words, they are apparently harmless.

If you do get a genome sequence that identifies all variants, you’re ahead. But knowing you have a variant doesn’t tell you what it means. That requires judgment. There are many millions of variants and many are rare or not understood. Even when we know a variant is tied to a disease, we may not know how much it increases risk. Labs and scientists disagree.

You might get a “polygenic” risk score that combines different common variants. These scores are most helpful for research because they describe broad tendencies toward disease rather than specific risk. They’re not likely, then, to help you much in predicting your own risk of disease (Horton et al., 2019).

Or you might be told that you have a well-known variant—for example, suggesting that you are more likely to develop Alzheimer’s disease or another illness.

Or you might learn that you carry a gene for a recessive genetic condition that could affect a future child if your partner is also a carrier. But you’d need more information to know how likely it is that your child will have this condition.

Some companies will give you access to your uninterpreted data. It’s up to you to get the data analyzed by a third party. Three big names are Promethease, LiveWello, and Interpretome.

You may get “false positives”

It’s standard for these tests to find risk. In one study, researchers recruited 199 patients of a large health maintenance organization aged 25 to 40 and offered them free testing of 15 variants associated with increased risk for diabetes, heart disease, high cholesterol, high blood pressure, lung cancer, colon cancer, skin cancer, or osteoporosis.

None of the participants had a current diagnosis for those problems. After testing, on average, they turned out to carry at least one variant associated with increased risk for six of the eight health conditions. This came to an average of nine risk-increasing variants out of the possible 15 (Kaphingst et al., 2012).

But that’s just one side of the equation. The chances are you have other protective factors that the test didn’t reveal.

If the condition hasn’t appeared in your family, you’re probably well-protected. Even if it has, your test result still doesn’t mean you will develop it. People often interpret their own risk as if they were being compared to someone with zero risk. In fact a screening test aims to tell you whether your risk is higher or lower than that of a particular population.

3D rendering of DNA structure
DNA is complex. Direct-to-consumer tests tend to overestimate our knowledge, and oversimplify genetic causality in disease.

A second way you might get a “false positive” is through plain old error. SNP-chip genotyping often turns up rare variants incorrectly—you’ll hear that you are at risk for a rare disease, but in fact you don’t have the variation it found (Horton et al., 2019).  

You may get a reassuring result that is incorrect

The information in a test may influence you to take fewer precautions. That’s a mistake. As one example, there are thousands of BRCA variants that increase the risk of breast cancer, but tests may check only for those common in people with Ashkenazi Jewish ancestry. If breast cancer runs in your family, you need more reliable results (Horton et al., 2019).

New information is being learned every day. Also, specialists disagree about the role of particular variants. A reassuring result from one test does not mean you are not at risk for a particular problem.

You could get very different answers depending on the DTC company you choose. One very basic area of disagreement among labs is how to use information about race, ethnicity or ancestry in predicting disease risk (Popejoy et al., 2020).   

If you have symptoms that you think may be related to the results in a DTC test, talk to your doctor and ask that the tests results be confirmed.

Is there any really reliable test?

The most helpful and reliable information may be an analysis of your potential response to particular drugs, including the blood-thinners clopidogrel (Plavix) and warfarin (Coumadin), which help prevent clots. In 2007, the Food and Drug Administration approved labeling for warfarin that suggests doctors use genetic tests to optimize dosing.

Will genetic testing improve? 

To speed these tests toward improved predictive power, the U.S. government helped create and fund a global database, ClinVar, which reported for the first time in the spring of 2015. Based on results contributed by university research projects and some private companies, the project is tracking some of the best-known variants. ClinVar has revealed more disagreement among the various contributors than scientists had expected (Harrison et al., 2017), for example about the use of ethnic and ancestry information in assessing disease risk (Popejoy et al., 2020).  

As more data is shared, some mysteries will be solved, but the process will take time.

Clinical labs have begun collaborating to resolve differences on how to interpret variants. For example, when four labs evaluated more than 6,000 variants submitted to ClinVar, they found their classifications – that is, their assessment of what risk that variant indicated - differed for 724 of them.

The four labs decided to tackle 242 of the variants and reached a consensus on 87 percent of that group. In another pilot project, nine laboratories compared their interpretations for 99 variants. At first, they agreed only on 34 percent. After sharing data, they reached agreement on 71 percent (Ray, 2021).

You can help in this effort to make the tests better by using labs that won a seal of approval from ClinVar  and by uploading your own de-identified genetic test reports to an online registry called Genome Connect.

Just remember, while genetic testing will improve, its progress doesn’t guarantee that you’ll get usable, accurate results for a small sum in a DTC test now or in the near future. This is one area where you still need to rely on physicians.  

 

Citations

Borry P, Bentzen HB, Budin-Ljøsne I, et al. The challenges of the expanded availability of genomic information: an agenda-setting paper. J Community Genet. 2018;9(2):103-116. doi:10.1007/s12687-017-0331-7  Published April, 2018.

Harrison, S., Dolinsky, J., Knight Johnson, A. et al. Clinical laboratories collaborate to resolve differences in variant interpretations submitted to ClinVar. Genet Med 19, 1096–1104 (2017). https://doi.org/10.1038/gim.2017.14 Published March, 2017.

Horton R, Crawford G, Freeman L, Fenwick A, Wright C F, Lucassen A et al. Direct-to-consumer genetic testing BMJ 2019; 367 :l5688 doi:10.1136/bmj.l5688 Published October 16, 2019.

Kaphingst KA, McBride CM, Wade C, et al. Patients' understanding of and responses to multiplex genetic susceptibility test results. Genet Med. 2012;14(7):681-687. doi:10.1038/gim.2012.22 Published July, 2012.

Popejoy AB, Crooks KR, Fullerton SM, et al. Clinical Genetics Lacks Standard Definitions and Protocols for the Collection and Use of Diversity Measures. Am J Hum Genet. 2020;107(1):72-82. doi:10.1016/j.ajhg.2020.05.005 Published July 2, 2020.

Ray T. ClinGen implementing strategies to Resolve variant Classification conflicts in clinvar. Genomeweb. https://www.genomeweb.com/molecular-diagnostics/clingen-implementing-strategies-resolve-variant-classification-conflicts?utm_source=TrendMD&utm_medium=TrendMD&utm_campaign=0&trendmd-shared=0#.YVTfy7hKiUk. Published June 22, 2021.

What happens in a genetics laboratory? EuroGentest. http://www.eurogentest.org/index.php?id=621. Published March 2011.

Zick CD, Mathews CJ, Roberts JS, Cook-Deegan R, Pokorski RJ, Green RC. Genetic testing for Alzheimer's disease and its impact on insurance purchasing behavior. Health Aff (Millwood). 2005;24(2):483-490. doi:10.1377/hlthaff.24.2.483 Published April, 2005.