40% false positive rate in DTC raw data
A landmark 2020 study in Genetics in Medicine sent 49 consumer genetic test results for clinical confirmation and found that 40% of variants reported in DTC raw data were false positives -- meaning the individual did not actually carry the variant. Always confirm consumer findings clinically before making health decisions. -- Tandy-Connor et al., Genetics in Medicine, 2020
The First Rule: Genetic Risk Is Not Destiny
Before you read a single result in your genetic health report, internalize this: for most common diseases, your genes are not your fate. They are one piece of a complex puzzle that includes environmental exposures, lifestyle factors, age, chance, and -- crucially -- medical care. A positive genetic risk finding means you should pay attention and possibly take preventive action. It does not mean you will develop the condition. A negative finding does not mean you are immune.
Only about 10% of cancers are strongly hereditary (driven by a single high-risk gene variant like BRCA1, BRCA2, or Lynch syndrome genes). The remaining 90% result from complex gene-environment interactions and acquired mutations during life. For cardiovascular disease, genetics explains roughly 40 to 60% of risk variance, with lifestyle factors (diet, exercise, smoking, stress) accounting for the rest. For most psychiatric and autoimmune conditions, the genetic contribution is even more diffuse.
Relative vs. Absolute Risk: The Most Common Misunderstanding
This distinction causes more confusion than any other concept in genetic reporting. When you open a genetic health report and see "2x increased risk," your instinct may be alarm. But you need to ask: 2x relative to what?
Consider this example: The general population has a 10% lifetime risk of developing type 2 diabetes. If your genetic profile indicates a 1.5x relative risk, your absolute risk becomes approximately 15% -- a 5-percentage-point increase, not 150%. The relative framing (50% increase) sounds dramatic. The absolute framing (5 percentage points) is sobering but manageable. Good genetic health reports present both numbers. Inferior reports emphasize relative risk alone.
Conversely, consider a condition with a very low baseline risk. If your relative risk for a rare cancer is 4x, but the population risk is 0.01% (1 in 10,000), your absolute risk becomes 0.04% (4 in 10,000) -- still very low. This is why understanding absolute risk matters: it prevents unnecessary anxiety over statistically impressive but clinically insignificant findings.
Polygenic Risk Scores: Promise and Limitations
Polygenic risk scores (PRS) represent one of the most exciting and controversial areas in genetic health reporting. Instead of looking at single high-impact variants (like BRCA mutations), PRS aggregates information from hundreds or thousands of common variants, each contributing a tiny amount to overall risk.
For some conditions, PRS provides clinically useful risk stratification. For coronary artery disease, individuals in the top percentile of PRS have a 3 to 4x higher lifetime risk than average -- comparable to having familial hypercholesterolemia, a well-known monogenic condition. For breast cancer, PRS combined with standard risk factors (family history, mammographic density) improves risk prediction over either alone.
However, PRS has three critical limitations:
- Ancestry bias: The vast majority of genome-wide association studies (GWAS) used to develop PRS have been conducted in European-ancestry populations. A 2022 study in Nature Genetics found that PRS predictive accuracy for type 2 diabetes was 4.5x lower in African-ancestry individuals and 2x lower in Hispanic-ancestry individuals compared to European-ancestry populations. If your ancestry is underrepresented in genetic research -- and this includes most non-European populations -- your PRS results are less reliable. -- Nature Genetics, 2022
- Condition-dependent accuracy: PRS works well for conditions with strong additive genetic architectures (coronary artery disease, type 2 diabetes, breast cancer). It works poorly for conditions dominated by rare variants, gene-environment interactions, or structural genomic changes (many neurodevelopmental disorders, autoimmune conditions).
- No clinical standards yet: Unlike monogenic testing, which follows well-established ACMG/AMP guidelines for variant classification, no universally accepted standards govern how PRS should be calculated, reported, or communicated to patients. Different companies using different algorithms and reference datasets may produce different risk scores for the same person.
Variants of Uncertain Significance (VUS): The Gray Zone
Perhaps the most frustrating result in any genetic health report is a variant of uncertain significance (VUS). You have a genetic change that has been detected, but science does not yet know whether it matters. VUS results are common -- 20 to 30% of variants found in comprehensive genetic testing receive this classification -- and they are almost always a source of anxiety for the recipient.
The critical rule with VUS findings: do not make medical decisions based on them. VUS means "we do not know." It does not mean "probably harmful" or "probably benign." A 2018 study in JAMA examined VUS reclassification rates over time and found that approximately 90% of VUS are ultimately downgraded to benign or likely benign as more data accumulates. The remaining ~10% are upgraded to likely pathogenic or pathogenic.
If you receive a VUS result, the appropriate action is periodic re-checking. Many clinical labs and some consumer providers automatically notify you when a VUS in your report is reclassified. Nebula Genomics and Dante Labs, for example, update their databases against the latest ClinVar, gnomAD, and medical literature several times per year.
Actionable vs. Non-Actionable Findings
The American College of Medical Genetics and Genomics (ACMG) maintains a list of 78 genes where incidental findings should be reported to patients because they are medically actionable -- meaning there are established interventions that reduce risk. This list includes:
- Cancer predisposition genes: BRCA1, BRCA2, TP53 (Li-Fraumeni), STK11 (Peutz-Jeghers), PTEN (Cowden), MLH1/MSH2/MSH6/PMS2 (Lynch syndrome), APC (familial adenomatous polyposis), and others
- Cardiovascular genes: MYBPC3, MYH7 (hypertrophic cardiomyopathy), KCNQ1, KCNH2, SCN5A (long QT syndrome), LDLR (familial hypercholesterolemia)
- Pharmacogenetic variants: Specific variants associated with severe adverse drug reactions (e.g., HLA-B*57:01 and abacavir hypersensitivity)
These are variants where knowing your status can change medical management -- earlier and more frequent cancer screening, prophylactic surgery, specific medications, or lifestyle modifications with proven benefit. By contrast, non-actionable findings include conditions with no established prevention, early detection, or treatment protocols. Receiving an APOE e4 result (associated with increased Alzheimer's risk on 23andMe), for example, provides risk information without a clear medical action plan.
The Importance of Genetic Counseling
If your genetic health report indicates a medically significant finding -- particularly a pathogenic or likely pathogenic variant in a cancer-predisposition or cardiovascular gene -- the single most important next step is consulting a genetic counselor. Genetic counselors are master's-level healthcare professionals trained to:
- Verify the finding (confirming that the reported variant is real and correctly classified)
- Interpret the finding in the context of your personal and family medical history
- Recommend appropriate medical screening, prevention, or treatment
- Help you communicate findings to at-risk family members
- Address the psychological impact of genetic risk information
Many consumer testing companies do not include genetic counseling in their base price. Some offer it as an add-on, while clinical tests through Invitae, Color Health, and hospital-based labs typically include it. If your consumer test finds something concerning, you can find a genetic counselor through the National Society of Genetic Counselors (nsgc.org) or ask your primary care physician for a referral.
Find the right DNA health test with proper interpretation
Best DNA Tests for HealthFrequently Asked Questions
What should I do if my consumer DNA test finds a cancer risk variant?
First, do not panic. Consumer tests can produce false positives -- the 2020 Genetics in Medicine study found 40% of DTC raw data variants were not real. Second, do not make medical decisions (including prophylactic surgery) based on a consumer test alone. The appropriate steps are: (1) save or print your report; (2) contact a genetic counselor or your primary care physician; (3) request clinical confirmation testing from a CLIA-certified lab, which will re-test the specific variant using a more accurate method; (4) only after clinical confirmation, work with your healthcare team to develop an appropriate screening or prevention plan. The cost of clinical confirmation testing is often covered by insurance if medically indicated.
How reliable are 23andMe's health risk reports?
23andMe's FDA-authorized health reports are analytically valid for the specific variants they test -- meaning the genotyping chip accurately detects the presence or absence of those variants. The problem is scope, not accuracy. 23andMe's BRCA report tests only 3 out of more than 4,000 known disease-associated BRCA1/BRCA2 variants. A negative 23andMe BRCA result does not rule out hereditary breast and ovarian cancer risk. Similarly, their late-onset Alzheimer's report tests only the APOE e4 variant; it does not capture other genetic or non-genetic risk factors. 23andMe's reports are accurate for what they test. What they test is a small fraction of clinically relevant genetic variation.
What is penetrance and why does it matter?
Penetrance is the probability that a person with a given genetic variant will develop the associated condition. High-penetrance variants, like those in BRCA1 (50 to 85% lifetime breast cancer risk), produce disease in a large proportion of carriers. Low-penetrance variants increase risk modestly but do not cause disease in most carriers. Most common genetic variants identified through GWAS have very low penetrance (less than 1.2x risk increase). Understanding penetrance is essential because it determines whether a genetic finding warrants aggressive preventive action (high penetrance) or simply increased awareness and screening (low penetrance). Clinical labs report penetrance data alongside variant classifications when available.